PER CURIAM Opinion; Concurrence by
Judge FERGUSON.ORDER
The per curiam opinion filed May 11, 2007 is hereby amended as follows:
1. At 485 F.3d at 516, line 15 of the slip opinion, the word “few” is changed to “number of.”
2. At 485 F.3d at 527, line 34 of the slip opinion, the word “few” is changed to “number of.”
With these amendments, Judges Ferguson, Reinhardt, and Thompson vote to deny the petition for rehearing.
Judge Reinhardt votes to deny the petition for rehearing en banc; Judges Ferguson and Thompson so recommend. The full court was advised of the suggestion for rehearing en banc. No judge requested a vote on whether to rehear the matter en banc. Fed. R.App. P. 35.
The petition for rehearing and petition for rehearing en banc are DENIED.
No future petitions for rehearing or rehearing en banc will be entertained.
AMENDED OPINION
PER CURIAM:In this pre-AEDPA habeas appeal, we confront once again the question whether defense counsel’s performance during the sentencing phase of a capital trial was so deficient as to violate the defendant’s right to counsel under the Sixth Amendment. In March of 1982, Appellant Joe Leonard Lambright and his co-defendant Robert Smith were convicted of first degree murder, kidnapping, and sexual assault. After a brief sentencing proceeding, Lambright was sentenced to death. His conviction and sentence were affirmed by the Arizona courts on direct appeal and in state post-conviction proceedings. Lambright then filed a habeas petition in federal district court seeking reversal on various grounds. The district court denied the petition, and Lambright appealed. In 1999, we reversed his conviction on the ground that the use of dual juries for a single trial of both Lambright and Smith violated due process. The en banc court then reheard the case, reversed the panel decision, and affirmed the denial of habeas relief with respect to the dual jury issue. After the case was returned to the panel for resolution of Lambright’s remaining claims, we rejected all of his contentions with the exception of penalty phase ineffective assistance of counsel. With respect to that issue, we remanded to the district court for an evidentiary hearing. After the evi-dentiary hearing, the district court ruled that Lambright’s trial counsel had not provided deficient representation, and further ruled that even if the performance was deficient, Lambright was not prejudiced thereby. Because we conclude that trial counsel’s performance was both deficient and prejudicial, we reverse and remand for issuance of a writ of habeas corpus and a new sentencing proceeding.
FACTUAL AND PROCEDURAL BACKGROUND
In March of 1980, Lambright and Smith were traveling across the country with Lambright’s girlfriend, Kathy Foreman. According to Foreman, who testified against Smith and Lambright in exchange for immunity, Smith complained to Lam-bright about the fact that he did not have a traveling companion with whom he could *1107engage in sexual relations. In response, Lambright said they would find him a girl. Lambright also said that “he would like to kill somebody just to see if he could do it.” While passing through the Tucson, Arizona area, the trio encountered a hitchhiker, Sandra Owen, and offered her a ride. Owen accepted, got in the car, and the four drove off. Smith raped Owen in the back seat of the car on the way to a mountain site where they all exited the vehicle. Smith then raped Owen a second time and thereafter began choking her. Lambright then stabbed Owen numerous times while Smith and Foreman restrained her. Finally, Lambright struck Owen in the head with a rock. Owen died as a result of her injuries.1
After the trio was arrested, Lambright admitted involvement in the offense, but told the police that Smith was the one who murdered Owen. Smith told police that Foreman and Lambright committed the murder. Foreman turned state’s evidence and testified that Lambright and Smith were the real culprits. On March 30,1982, both Lambright and Smith were convicted of first degree murder, sexual assault, and kidnapping. The state sought and obtained the death penalty for both.
I. Trial Counsel’s Penalty Phase Investigation
At both the guilt and penalty phases of his trial, Lambright was represented by attorney Carmine Brogna. Between Lam-bright’s conviction and the sentencing hearing, Brogna met with Lambright once for a little over an hour and spoke with him once briefly over the phone. He spent less than an hour reviewing the pre-sen-tence report and speaking with Lam-bright’s probation officer, and a total of three hours drafting and dictating the brief sentencing memorandum he submitted to the court and the subpoena he used for the appearance of his sole penalty phase witness, a guard at the jail in which Lambright was being held.
Although prior to sentencing Brogna became aware of Lambright’s long history of mental health problems, his two prior suicide attempts, and his resultant hospitalization in a psychiatric facility, Brogna did not discuss these matters with Lam-bright’s friends or family members, nor did he request Lambright’s medical or hospital records. Moreover, although he knew that Lambright had discussed traumatic combat experiences in Vietnam with the probation officer assigned to prepare a pre-sentence report and with the Pima County Court’s psychologist, he did not attempt to obtain any information about Lambright’s experiences in Vietnam nor their effect on him. Even after the court’s psychologist drafted a report in which he concluded that Lambright suffered from antisocial personality disorder, Brogna did not contact the psychologist to discuss this diagnosis, nor did he attempt to have another psychologist or psychiatrist evaluate Lambright.
Brogna also knew that Lambright had a serious drug problem, but did no investigation with respect to the extent of his drug use or its effect on his mental state or behavior. He likewise did not seek to obtain Lambright’s school or military records. Nor did he contact either of Lam-bright’s former wives. Although Brogna traveled with counsel for Lambright’s co-defendant to Texas and Louisiana for five days and spoke with a number of potential witnesses prior to the guilt phase of Lam-bright’s trial, the trip did not yield any useful mitigating evidence, and Brogna did *1108no additional investigating thereafter. It is unclear whether Brogna sought information for use at the penalty phase proceeding from any of the people he spoke with during this trip. Although Brogna did speak to Lambright’s sister when he was in Texas, he asked her primarily about Lambright’s conduct around the time of the crime, and he does not recall whether he even raised the issues of Lambright’s upbringing, mental health, drug abuse, or any other potential mitigating factors.
II. Presentation of Mitigating and Aggravating Evidence
The sum total of the mitigating evidence that Brogna offered at sentencing comprises less than three pages of a double-spaced transcript, and all of it related solely to Lambright’s conduct in jail. Although, as noted above, Brogna knew that there were indications that Lambright was mentally ill, Brogna failed to present any mitigating psychiatric or psychological testimony. Nor did he call any of Lambright’s family members or friends to testify about his unfortunate childhood, his history of mental instability, his suicide attempts, or his long-term drug abuse. Instead, Brogna promised the court that he had “one witness who will be very brief.” The sole witness called by Brogna was George Din-nen, an officer at the Pima County Jail, who testified that in the six months in which he had supervised Lambright, there had been no “problems with him” and that he had “always been respectful, courteous, and cooperative.” In addition to calling this witness, Brogna submitted a two-and-a-half-page legal memorandum listing various potentially mitigating circumstances, including (1) the fact that a deal was cut with Lambright’s accomplice, Kathy Foreman, in exchange for her cooperation with the prosecution, (2) Lambright’s “lack of a prior violent criminal record,” (3) “the Defendant’s record while incarcerated,” (4) that Lambright “is amenable to some type of treatment,” (5) the fact that his “military record is a good one,” and (6) that “[t]here is no record indicating he was a ‘juvenile delinquent.’ ” None of these potentially mitigating circumstances was described with any particularity, nor did Brogna point to any evidence in the record to support the assertions made in his memorandum.
In addition to the scant mitigating evidence alluded to by Brogna, the sentencing judge considered the pre-sentence report prepared by the probation officer assigned to the case, a report written by Dr. Richard Hinton — a clinical psychologist with the Pima County Court’s clinic — and the testimony of the prosecution’s sole penalty phase witness, who was the lead detective on the case, and who testified about the circumstances under which the victim was raped and murdered.
Dr. Hinton’s report was based on a psychological evaluation of Lambright, which he performed shortly after Lambright’s conviction. Brogna did not provide Dr. Hinton with any information regarding Lambright’s childhood, his service in the military, his history of mental health problems, his medical history, or his long-term drug abuse, either prior to the time that Dr. Hinton conducted his evaluation or prior to the time that he prepared his report. As a result, Dr. Hinton relied exclusively on his interview with Lam-bright and on police reports, newspaper articles, and other materials related to the offense.
Dr. Hinton’s report, which presented a very negative and damning picture of Lambright, contained some basic information about potentially mitigating aspects of Lambright’s background, including the fact that Lambright had moved frequently as a child, that he had served in the Air Force in Vietnam during which time he had allegedly engaged in hand-to-hand *1109combat, that he had once suffered a nervous breakdown which necessitated his admission to the psychiatric unit at a Veterans Administration hospital in Houston, Texas, that he attempted suicide on two occasions, and that he had a long history of drug abuse. Based on his interview with Lambright, Dr. Hinton diagnosed him with antisocial personality disorder and included this diagnosis in his report.
The pre-sentence report, like Dr. Hinton’s report, contained some potentially mitigating information regarding Lam-bright’s personal history but was overwhelmingly more aggravating than mitigating. It contained a detailed description of the offense, a summary of a telephone conversation between the victim’s mother and the probation officer who prepared the report, in which the victim’s mother stated that both defendants should receive the death penalty. It also contained a summary of the defendant’s social history, emphasizing his negative personality traits and behavior. The pre-sentence report did note, however, that “[t]he defendant has no official, verifiable record of juvenile offenses,” that Lambright had been raised by “a very strict, hypochondriacal, Pentecostal mother,” and that “[a]t the age of nine or ten, he began running away from home.”
Regarding Lambright’s history of mental health problems, the pre-sentence report contained the following information:
[T]he defendant described how he had a breakdown after his return from Vietnam, apparently experiencing generalized feelings of paranoia in conjunction with feelings of failure over his first marriage. His sister described how they found him cowering on their front porch one evening, holding a butcher knife to protect himself from imagined attackers.
After spending three days in a Houston veterans administration hospital, he left against medical advice. Attending physicians apparently told his sister that if the defendant did not receive long-term therapy, he would probably spend the rest of his life running from place to place ...
As for Lambright’s history of drug use, the pre-sentence report noted that “[t]he defendant has used and abused various substances over the past ten years, most notably marijuana, amphetamines, and alcohol .... He has never received any type of treatment for substance abuse.”
Brogna’s argument at the close of the sentencing proceeding, like his presentation of mitigating evidence, was extremely brief. He asked the court to consider the fact that Kathy Foreman, although a direct participant in the crime, received no punishment, stated that Dr. Hinton’s conclusion that Lambright suffers from antisocial personality disorder was unfounded, and asked the court to consider Lam-bright’s lack of a violent history. He did not ask for leniency, but instead, without any explanation of his reasons, merely requested that the court impose a sentence of life in prison.
III. The Death Verdict and State Appeals
After the very brief sentencing proceeding, the sentencing judge found that only one aggravating circumstance existed, namely “that the offense was committed in an especially heinous, cruel or depraved manner.” He nonetheless sentenced Lam-bright to death because he found that the limited mitigating evidence presented was insufficiently substantial to outweigh this single aggravating factor. The judge noted that under Arizona law, “ ‘[t]he Court ... shall impose a sentence of death if it finds one or more aggravating circumstances and finds no mitigating circumstances sufficiently substantial to call for *1110leniency.’ ” In light of his finding that the mitigating evidence presented was insubstantial, he concluded that “the law demands the maximum penalty ...” The sentencing judge did not list Lambright’s long history of substance dependency or his mental health problems among the mitigating factors he considered. With respect to Lambright’s childhood, he said only that Lambright’s “unsettled early life” was not significantly mitigating. Lambright’s conviction and death sentence were affirmed by the Arizona Supreme Court both on direct appeal and in subsequent state habeas proceedings. State v. Lambright, 138 Ariz. 63, 673 P.2d 1 (1983). Thereafter, in April of 1987, Lambright petitioned for federal habeas corpus relief on various grounds.
IV. Federal Habeas Review
In July of 1996, the district court denied Lambright’s habeas petition, and Lam-bright appealed. In January of 1999, we issued an opinion reversing his conviction on the ground that the “single trial before ‘dual juries’ ” violated his right to due process under the Fourteenth Amendment. Lambright v. Stewart, 167 F.3d 477, 479 (Lambright /), withdrawn, 177 F.3d 901 (9th Cir.1999). The court later reheard the case en banc, reversed the panel’s decision, and affirmed the district court’s denial of relief with respect to the dual jury issue. Lambright v. Stewart, 191 F.3d 1181, 1182, 1187(9th Cir.1999) (en banc) (Lambright //). The case was then returned to this panel for resolution of Lambright’s remaining claims. Id. at 1187. We affirmed the district court’s denial of his habeas petition with respect to all of his guilt phase claims and all but one of his penalty phase claims, but reversed the district court’s ruling that his penalty phase ineffective assistance of counsel claim had been procedurally defaulted. Lambright v. Stewart, 241 F.3d 1201, 1203 (9th Cir.2001) (Lambright III); Lambright v. Stewart, 5 Fed.Appx. 712, 713-15 (9th Cir.2001). After finding that Lambright “presented more than enough evidence to establish a colorable claim for deficient performanee[,]” we remanded the case to the district court for an evidentiary hearing on the issue of “whether Lambright was denied effective assistance of counsel at sentencing because of the failure to investigate and present evidence of his psychiatric condition and social history.” Lambright III, 241 F.3d at 1208.
V. The Evidentiary Hearing
In November of 2003, the district court held a six-day evidentiary hearing on the issue of ineffective assistance of counsel. The testimony, affidavits, and other evidence introduced at the evidentiary hearing revealed that there was a substantial amount of mitigating evidence that Brog-na could have investigated, developed, and presented at Lambright’s sentencing hearing. With respect to Lambright’s childhood, the evidence presented at the evidentiary hearing revealed that Lam-bright’s mother beat him severely and regularly from the time he was a small child until he was fourteen years old, usually by whipping him with various objects, kicking him, or striking him.
Testimony at the hearing also revealed that Lambright’s mother presented symptoms of severe hypochondria, and was profoundly addicted to prescription drugs throughout Lambright’s childhood. She took sleeping pills, Valium, and Darvon on a daily basis and, as a result, spent most of her time in bed, leaving Lambright to fend for himself. Starting around the time that he was nine years old, she often forced him to take sedatives in order to make him settle down or sleep. She was also extremely verbally and emotionally abusive toward Lambright throughout his childhood. Lambright’s father ignored his *1111wife’s abusive behavior and did nothing to stop it.
Evidence was also presented regarding the poverty that Lambright experienced as a child. According to Lambright’s sister, the family was “very poor” throughout their childhood. They often lived in homes with no running water or indoor plumbing. Once, they were forced to move into a rat-infested house in which the walls and ceilings were lined with cardboard to block holes. Moreover, throughout Lambright’s childhood, his family moved every six to nine months because his father had difficulty maintaining steady employment. As a result, Lambright never attended the same school for more than a year, had difficulty forming relationships with classmates, and was a frequent target of bullies. He stopped attending school altogether after the ninth grade.
Evidence related to Lambright’s mental health problems and long history of drug abuse was also presented. Lambright’s first wife testified that, prior to 1968 when Lambright was deployed to Vietnam, he was a kind, gentle, and loving person who neither drank nor did drugs. When he returned from Vietnam, he seemed deeply traumatized and his personality had radically changed. He had trouble maintaining stable, healthy relationships, was constantly jumpy and agitated, had difficulty sleeping, was plagued by nightmares, and became deeply depressed. He would often lock himself in another room and would say that he wished he were dead. He eventually became so despondent that he attempted suicide by driving his car into a tree.
On one occasion, Lambright’s sister awoke to find him pounding on her door at two o’clock in the morning. When she came to the door, she saw Lambright cowering against a wall, swinging a butcher knife. He was hallucinating and appeared to be hiding from an imagined attacker. He kept saying “I gotta get to ’em. I see ’em. They’re burning Mama and Daddy, they’re burning.” During this incident Lambright was crying, incoherent, and “shaking all over.”
Thereafter, Lambright was admitted to the mental health unit of the Houston VA hospital. After three days of treatment, Lambright left the hospital against the advice of his doctors because he was concerned about losing his job and not being able to support his wife and young son.
In 1970, Lambright began working as a truck driver and started using speed and diet pills to stay awake for long periods of time. In 1971, after his father’s death, he became profoundly depressed and began drinking heavily and regularly. In 1972, Lambright began using marijuana and experimenting with various other drugs, including acid, hash, cocaine, and mushrooms. In 1974, Lambright attempted suicide for a second time, this time by intentionally overdosing on drugs, and was briefly hospitalized as a result. Lam-bright began using crystal methamphetamine (“meth”) in 1976 or 1977, and by 1979, he would stay high on meth for weeks at a time.
With respect to his time in Vietnam, Lambright testified about a traumatic combat experience in which he witnessed the violent death of a close friend who was shot and killed by the Viet Cong. However, the testimony of other witnesses, as well as other evidence presented at the evidentiary hearing, indicated that this event did not occur. Conceivably, it was a figment of Lambright’s imagination. Also conceivably, it was a story he concocted.
Expert testimony was also presented with respect to Lambright’s mental health and substance abuse problems. The psychiatrist who testified on Lambright’s behalf, Dr. Barry Morenz, concluded that Lambright’s Vietnam experience, along *1112with the abuse Lambright experienced at the hands of his mother, caused Lambright to develop post-traumatic stress disorder (“PTSD”). The psychologist testifying on behalf of the state, Dr. Gina Lang, rejected this diagnosis, primarily because she believed that Lambright’s account of his experience in Vietnam was fabricated. Both experts agreed, however, that Lam-bright suffers from a depressive disorder and from polysubstance dependency. Dr. Lang also concluded that Lambright suffers from a personality disorder with antisocial, borderline, and inadequate features.
Dr. Hinton, the psychologist who submitted a report to the sentencing judge in 1982 stating that Lambright suffered from antisocial personality disorder, also testified at the evidentiary hearing. He initially stated that if he had been provided by Brogna with information about Lam-bright’s background, he might have been inclined to infer that Lambright suffered from PTSD, but upon cross-examination acknowledged that such a diagnosis would be improbable. He did not deny, however, that Lambright suffered from some type of mental disorder.
VI. The District Court’s Memorandum of Decision and Order
On August 30, 2004, the district court issued a memorandum of decision and order denying Lambright’s habeas petition. The court found that “Petitioner’s testimony with respect to the trauma underlying his expert’s PTSD diagnosis (the alleged combat experience) is less than credible.” The court further found Dr. Morenz’s PTSD diagnosis unpersuasive because it “rests almost exclusively on the alleged combat experience as the traumatic trigger event,” and instead found Dr. Lang’s assessment of Lambright’s mental health problems to be the more persuasive of the two. Moreover, the court stated that even if it were “persuaded that Petitioner suffered from PTSD, Petitioner has failed to show how the disorder affected his criminal behavior.” The court further found “no credible evidence that methamphetamine abuse contributed to Petitioner’s criminal actions” and that “Petitioner ... presented no evidence to support the assertion that he was under the influence of drugs at the time of the crime.”
Analyzing Lambright’s claim of ineffective assistance of counsel, the district court concluded that Brogna had provided adequate representation in spite of the fact that he failed to present mitigating evidence related to Lambright’s prolonged drug abuse, mental health problems, and abusive childhood. With regard to the drug abuse, the court stated that it “cannot say that it was unreasonable not to focus on drug abuse as a mitigating factor in the absence of any explanatory nexus to the crime.” Turning to Brogna’s failure to present evidence related to Lambright’s history of mental instability and specifically his experiences in Vietnam, the court stated that
[cjounsel cannot be faulted for failing to investigate PTSD when neither Petitioner, his sister nor Dr. Hinton suggested that the few months Petitioner spent in Vietnam so altered him that the experience provided a possible explanation for Petitioner’s participation in the killing.
The court made no further comment regarding Lambright’s mental instability or the expert testimony supporting the existence of a mental disorder other than PTSD. With regard to Lambright’s abusive childhood, the court concluded that “Petitioner’s upbringing was not so horrific that it significantly affected Petitioner’s conduct at the time of the crime.” Accordingly, the court held that “Brogna’s presentation [of mitigating evidence] was not constitutionally deficient.”
The court further held that, “[ajssuming Brogna’s representation was objectively *1113unreasonable, Petitioner is unable to establish that Brogna’s errors prejudiced him.”- Specifically, the court concluded that
[bjecause Petitioner has failed to establish that he was in combat or suffered from PTSD, the Court cannot say that Petitioner’s sentence would have been different had Brogna conducted a full-blown investigation into Petitioner’s Vietnam experiences. Moreover, even if PTSD were a viable diagnosis, it has not been shown to have any explanatory or exculpatory attributes with respect to [the victim’s] murder ...
Regarding Lambright’s substance abuse problem, the district court concluded that
the sentencing judge was already well aware of it through the presentence report .... In addition, Petitioner never argued that he was under the influence of alcohol or drugs when he committed the murder.
Therefore, any evidence of substance abuse could not have mitigated the circumstances of the crime itself.
With respect to Lambright’s childhood, the court concluded that “Petitioner’s childhood, though bleak, was largely identified to the sentencing judge.... The Court has already noted that Petitioner failed to prove his upbringing affected his ability to know right from wrong or to control his conduct.” Based on these conclusions, the court held that because “the avenues of potential mitigation evidence are either unsubstantiated or were already before the sentencing judge, the Court is not persuaded that any deficiencies in counsel’s representation rendered the sentencing proceeding ‘fundamentally unfair or unreliable.’ ”
On appeal, Lambright argues that the district court erred in (1) holding that Brogna’s investigation and presentation of mitigating evidence was not constitutionally deficient, and (2) holding that any deficiency in Brogna’s performance was not prejudicial. Lambright also argues that the district court committed two specific errors in reaching these conclusions: (1) that it erroneously required a causal nexus between his proffered mitigating evidence and the crime, and (2) that it erred in concluding that Lambright could not assert, during cross-examination, a Fifth Amendment privilege with regard to questions related to the offense and likewise erred in drawing a negative inference regarding certain contested factual issues based on his refusal to answer such questions.2
DISCUSSION
I. Standard of Review
Lambright’s federal habeas petition was filed prior to the enactment of the Antiterrorism and Effective Death Penalty Act of 1996 (“AEDPA”), and thus, pre-AEDPA standards apply to his claims. Douglas v. Woodford, 316 F.3d 1079, 1085 (9th Cir.2003). We review the district court’s decision to deny habeas relief de *1114novo. Raley v. Ylst, 470 F.3d 792, 799 (9th Cir.2006). “Under pre-AEDPA law, we consider a claim alleging ineffective assistance of counsel as a mixed question of law and fact that we review de novo.” Summerlin v. Schriro, 427 F.3d 623, 628 (9th Cir.2005) (en banc) (citing Rios v. Rocha, 299 F.3d 796, 799 n. 4 (9th Cir.2002)). “We review for clear error, however, the district court’s findings of fact.” Frierson v. Woodford, 463 F.3d 982, 988 (9th Cir.2006). Finally, “[b]ecause this is a pre-AEDPA case, we do not review the state court’s legal conclusions to determine whether they are ‘objectively unreasonable;’ rather, we ‘simply resolve the legal issue on the merits, under the ordinary rules.’ ” Summerlin, 427 F.3d at 628(quoting Belmontes v. Brown, 414 F.3d 1094, 1101 (9th Cir.2005), rev’d on other grounds, Ayers v. Belmontes, — U.S. -, 127 S.Ct. 469, 166 L.Ed.2d 334 (2006)).
II. The District Court’s Erroneous Application of a Nexus Requirement
At the outset, we reject the district court’s analysis with respect to deficient performance and prejudice because it is predicated on the application of a legal test that is clearly inapplicable. The district court disregarded virtually all of the mitigating evidence that Lambright presented at the evidentiary hearing on the basis that it had no “explanatory nexus” to the crime. In so doing, it misapplied controlling Supreme Court and Ninth Circuit precedent. In Tennard v. Dretke, 542 U.S. 274, 124 S.Ct. 2562, 159 L.Ed.2d 384 (2004), the Supreme Court explicitly rejected the Fifth Circuit’s requirement that mitigating evidence have some nexus or causal connection to the crime of which the capital defendant was convicted in order for a jury’s failure to consider that evidence to be deemed prejudicial. Id. at 289, 124 S.Ct. 2562. The Court held that
[t]he Fifth Circuit’s test has no foundation in the decisions of this Court.... When we [have] addressed directly the relevance standard applicable to mitigating evidence in capital cases ... we[have] spoke[n] in the most expansive terms.... The Fifth Circuit’s test is inconsistent with these principles. Most obviously, the test will screen out any positive aspect of a defendant’s character because good character traits are ... no[t] typically traits to which criminal activity is “attributable.”
Id. at 284-85, 124 S.Ct. 2562. Turning specifically to the “nexus” element of the Fifth Circuit’s test, the Court held that
[t]he Fifth Circuit was ... wrong to have refused to consider ... [Tennard’s claim] on the ground that Tennard had not adduced evidence that his crime was attributable to his low IQ. In Atkins v. Virginia, 536 U.S. 304, 316[, 122 S.Ct. 2242, 153 L.Ed.2d 335] (2002), we explained that impaired intellectual functioning is inherently mitigating ... Nothing in our opinion suggested that a mentally retarded individual must establish a nexus between her mental capacity and her crime before the Eighth Amendment prohibition on executing her is triggered. Equally, we cannot countenance the suggestion that low IQ evidence is not relevant mitigating evidence ... unless the defendant also establishes a nexus to the crime.
Id. at 287, 124 S.Ct. 2562.
In Smith v. Texas, 543 U.S. 37, 45, 125 S.Ct. 400, 160 L.Ed.2d 303 (2004), the Court reiterated that it had “rejected the ... ‘nexus’ requirement in Tennard,” and stated that “petitioner’s evidence [regarding his troubled childhood and limited mental abilities] was relevant for mitigation purposes is plain under our precedents.” Id. (citing Penry v. Lynaugh, 492 U.S. 302, 319-20, 109 S.Ct. 2934, 106 *1115L.Ed.2d 256 (1989); Payne v. Tennessee, 501 U.S. 808, 822, 111 S.Ct. 2597, 115 L.Ed.2d 720 (1991); Boyde v. California, 494 U.S. 370, 377-78, 110 S.Ct. 1190, 108 L.Ed.2d 316 (1990); Eddings v. Oklahoma, 455 U.S. 104, 114, 102 S.Ct. 869, 71 L.Ed.2d 1 (1982)). Indeed, the Court has characterized the “nexus” test as “a test we never countenanced and now have unequivocally rejected.” Id.
We, too, have addressed the question whether evidence with no explanatory nexus to the crime at issue should be considered mitigating in Smith v. Stewart, 140 F.3d 1263, 1271 (9th Cir.1998), and we did so in the context of assessing whether counsel’s deficient performance in a capital sentencing proceeding was prejudicial. Id. We held specifically that the issue of whether counsel’s failure to present evidence related to antisocial personality disorder and long-term drug use was prejudicial did not turn on “whether those precluded [the defendant] from knowing right from wrong, as the post-sentencing court and the district court seemed to have assumed” and as they would have to have done for there to be a nexus between those factors and the crime. Id. Instead, we held that “they need not do so in order to be ... miti-gators.” Id.
The reason for rejecting a nexus requirement is clear: “ ‘the fundamental respect for humanity underlying the Eighth Amendment ... requires consideration of the character and record of the individual offender ... as a constitutionally indispensable part of the process of inflicting the penalty of death.’ ” Eddings, 455 U.S. at 112, 102 S.Ct. 869(quoting Woodson v. North Carolina, 428 U.S. 280, 304, 96 S.Ct. 2978, 49 L.Ed.2d 944 (1976)). If evidence relating to life circumstances with no causal relationship to the crime were to be eliminated, significant aspects of a defendant’s disadvantaged background, emotional and mental problems, and adverse history, as well as his positive character traits, would not be considered, even though some of these factors, both positive and negative, might cause a sentencer to determine that a life sentence, rather than death at the hands of the state, is the appropriate punishment for the particular defendant. This is simply unacceptable in any capital sentencing proceeding, given that “treating each defendant in a capital case with that degree of respect due the uniqueness of the individual,” and determining whether or not he is deserving of execution only after taking his unique life circumstances, disabilities, and traits into account, is constitutionally required. Lockett v. Ohio, 438 U.S. 586, 605, 98 S.Ct. 2954, 57 L.Ed.2d 973 (1978).
Here, the district court relied heavily on its finding that Lambright had not shown a nexus between his proffered mitigating evidence and the crime, flatly rejecting the majority of the mitigating evidence he offered on that basis. Indeed, it is apparent from the district court’s order that it either did not consider mitigating any evidence without an explicit nexus to the crime, or that it gave such evidence de minimus weight. Because the district court’s rejection of Lambright’s mitigating evidence on that basis violates the rule set forth in Tennard, Smith v. Texas, Smith v. Stewart, and their predecessors, we hold that its analysis of Lambright’s ineffective assistance of counsel claim was fundamentally flawed.
III. Ineffective Assistance of Counsel
We begin by assessing whether Brog-na’s performance at sentencing was so deficient as to violate Lambright’s right to counsel under the Sixth Amendment. “The Sixth Amendment right to counsel in a criminal trial includes ‘the right to the effective assistance of counsel.’ ” Sum-*1116merlin, 427 F.3d at 629 (quoting McMann v. Richardson, 397 U.S. 759, 771 n. 14, 90 S.Ct. 1441, 25 L.Ed.2d 763 (1970)). “This right extends to ‘all critical stages of the criminal process,’ including capital sentencing.” Id. (citations omitted).
In order to prevail on a claim of ineffective assistance of counsel, a petitioner must show that (1) his trial counsel’s performance “fell below an objective standard of reasonableness”; and (2) “there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Strickland v. Washington, 466 U.S. 668, 687-88, 694, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984).
A. Deficient Performance
Under Strickland, counsel’s competence is presumed and thus Lambright must rebut this presumption by demonstrating that Brogna’s performance was unreasonable under prevailing professional norms and was not the product of sound strategy. See id. at 688-89, 104 S.Ct. 2052. Judicial scrutiny of counsel’s performance is highly deferential, and thus we must evaluate Brogna’s conduct from his perspective at the time it occurred, without the benefit of hindsight. Id. at 689, 104 S.Ct. 2052. “[Strategic choices made after thorough investigation of [the relevant] law and facts relevant to plausible options are virtually unchallengeable.” Id. at 690, 104 S.Ct. 2052. However,
strategic choices made after less than complete investigation are reasonable precisely to the extent that reasonable professional judgments support the limitations on investigation. In other words, counsel has a duty to make reasonable investigations or to make a reasonable decision that makes particular investigations unnecessary. In any ineffectiveness case, a particular decision not to investigate must be directly assessed for reasonableness in all the circumstances ...
Id. at 690-91, 104 S.Ct. 2052; see also Wiggins v. Smith, 539 U.S. 510, 521, 123 S.Ct. 2527, 156 L.Ed.2d 471 (2003) (quoting Strickland, 466 U.S. at 690-91, 104 S.Ct. 2052). Similarly, a decision not to present a particular defense or not to offer particular mitigating evidence is unreasonable unless counsel has explored the issue sufficiently to discover the facts that might be relevant to his making an informed decision. Wiggins, 539 U.S. at 522-23, 123 S.Ct. 2527; Stankewitz v. Woodford, 365 F.3d 706, 719 (9th Cir.2004).
The Supreme Court has “declined to articulate specific guidelines for appropriate attorney conduct and instead ha[s] emphasized that ‘the proper measure of attorney performance remains simply reasonableness under prevailing professional norms.’” Wiggins, 539 U.S. at 521, 123 S.Ct. 2527(quoting Strickland, 466 U.S. at 688, 104 S.Ct. 2052). “However, general principles have emerged regarding the duties of criminal defense attorneys that inform our view as to the ‘objective standard of reasonableness’ by which we assess attorney performance, particularly with respect to the duty to investigate.” Summerlin, 427 F.3d at 629.
For example, we have held that “ ‘[t]o perform effectively ... counsel must conduct sufficient investigation and engage in sufficient preparation to be able to “present[] and explain[] the significance of all the available [mitigating] evidence.” ’ ” Allen v. Woodford, 395 F.3d 979, 1000(9th Cir.2005) (citing Mayfield v. Woodford, 270 F.3d 915, 927 (9th Cir.2001) (en banc)); see also Summerlin, 427 F.3d at 630. Indeed, we have consistently held that “ ‘it is imperative that all relevant mitigating information be unearthed for consideration at the capital sentencing phase.’ ” Wallace v. Stewart, 184 F.3d *11171112, 1117 (9th Cir.1999) (quoting Caro v. Calderon, 165 F.3d 1223, 1227 (9th Cir.1999)).
“To that end, the investigation should include inquiries into social background and evidence of family abuse.” Summerlin, 427 F.3d at 630. Attorneys representing defendants in capital sentencing proceedings also have a “ ‘duty to investigate and present mitigating evidence of mental impairment’ ...[,] [which] includes examination of mental health records.” Id. (quoting Bean v. Calderon, 163 F.3d 1073, 1080(9th Cir.1998) and citing Deutscher v. Whitley, 884 F.2d 1152, 1161 (9th Cir.1989)); see also Caro v. Woodford, 280 F.3d 1247, 1254 (9th Cir.2002). Furthermore, “counsel has an affirmative duty to provide mental health experts with information needed to develop an accurate profile of the defendant’s mental health.” Caro v. Woodford, 280 F.3d at 1254. “The defendant’s history of drug and alcohol abuse should also be investigated.” Summerlin, 427 F.3d at 630(citing Jennings v. Woodford, 290 F.3d 1006, 1016-17 (9th Cir.2002)). Lambright’s counsel did almost none of this in the present case.
Moreover, when “tantalizing indications in the record” suggest that certain mitigating evidence may be available, those leads must be pursued. Stankewitz, 365 F.3d at 719-20; see also Summerlin, 427 F.3d at 632(finding ineffective assistance where trial counsel failed to obtain readily available evidence concerning possible mental health mitigation where his chent’s prior attorney told him there were indications that the defendant was mentally ill); Wallace, 184 F.3d at 1115— 16 (finding ineffective assistance where trial counsel spent only two hours interviewing potential witnesses, just over thirty minutes with a psychiatric expert, and failed to contact known and willing witnesses); Seidel v. Merkle, 146 F.3d 750, 756 (9th Cir.1998) (finding ineffective assistance where defense counsel “failed to conduct even a minimal investigation in order to make an informed decision” regarding his client’s mental health defense).
In Correll v. Ryan, 465 F.3d 1006 (9th Cir.2006), for example, we held that petitioner’s trial counsel had been ineffective in failing to adequately investigate and present evidence related to the defendant’s background, mental health problems, and long-term drug addiction at the penalty phase of his capital trial. Id. at 1012-15. We noted that Correll’s
attorney knew that, among other things, Correll came from a dysfunctional family, sustained a serious head injury, was committed to various psychiatric facilities, and that he was addicted to drugs; yet defense counsel did not obtain the records nor did he interview witnesses concerning these matters. Counsel did meet with the family members who would cooperate, but he admitted that he met only once with Correll’s father, sister, and brother ... and probably spent “[a] couple hours” with them. Counsel did not obtain Correll’s school records ... [He] did not obtain Correll’s medical records.... [He] could not recall what efforts he made to gather Correll’s psychiatric records, although defense counsel did remember that he did not obtain records from Correll’s stays at various mental health centers.
Id. at 1011. Assessing counsel’s mitigation investigation under the standards set forth above, we concluded that
in light of the abundance of classic mitigation evidence of which counsel was aware, his almost complete failure to investigate is startling.... Defense counsel’s failure to investigate falls far short of any objectively reasonable standard against which we might measure attorney performance under the standards of the Sixth Amendment.
*1118Id. at 1011-12. Turning to counsel’s presentation of evidence at the penalty phase, we observed that “[a]s anemic as the defense counsel’s investigation was, his presentation of mitigating evidence ... was [even] worse.” Id. at 1012. Defense counsel “did not call a single witness to testify. He did not introduce any evidence. Rather, he elected to allow the judge to make a decision on whether to sentence Correll to death based solely on the state’s evidence and the pre-sentence report.” Id. Moreover, “[d]efense counsel’s entire mitigation argument was contained in less than one page of a sentencing memorandum[,] ... [and t]he entirety of his oral argument at the penalty phase consisted] of approximately 7 pages of transcript.” Id. at 1013.
The facts of the case at hand and those in Cowell are strikingly similar. Here, as in Correll, Brogna failed to do even a minimal investigation of “classic mitigation evidence,” notwithstanding the fact that he knew such evidence potentially existed. Specifically, although prior to the imposition of the capital sentence Brogna became aware of Lambright’s history of mental health problems, the fact that he had attempted suicide on two prior occasions, and that he had been hospitalized in a psychiatric facility as a result, Brogna did not discuss these issues with Lam-bright’s friends or family members, nor did he request Lambright’s medical or hospital records. Although he knew that Lambright had discussed traumatic experiences in Vietnam with the probation officer assigned to prepare the pre-sentence report and with the Pima County Court’s psychologist, he did not attempt to obtain any information about these experiences or their effect on him. Even after the court’s psychologist prepared a report in which he concluded that Lambright suffered from antisocial personality disorder, Brogna did not contact the psychologist to discuss this diagnosis, nor did he attempt to have another psychologist or psychiatrist evaluate Lambright. He likewise failed to hire an independent investigator, despite his awareness that the court would pay for these additional expenses.
Brogna also knew that Lambright had a serious drug problem, but he conducted no investigation with regard to the extent of his drug use or its effect on his general mental state or behavior. He likewise did not seek to obtain Lambright’s school or military records. Nor did he contact either of Lambright’s former wives. Although Brogna traveled with counsel for Lambright’s co-defendant to Texas and Louisiana for five days and spoke with a number of potential witnesses prior to the guilt phase of Lambright’s trial, the trip did not yield any useful mitigating evidence, and Brogna did no additional investigating thereafter. Moreover, it is unclear whether any of the individuals he spoke with during this trip were people he viewed as potential penalty phase witnesses. When Brogna spoke to Lam-bright’s sister, he primarily asked her about Lambright’s conduct around the time of the crime and does not appear to have asked her about Lambright’s upbringing, his mental health, or other potential mitigating evidence.3 Accordingly, *1119as in Correll, Brogna’s investigation of potentially mitigating evidence was utterly deficient under the standards established by this court and by the Supreme Court.
In light of his woefully inadequate investigation, it is not surprising that Brogna’s presentation at the sentencing proceeding was minimal and markedly uncompelling. He spent only five and a half hours obtaining evidence and preparing for the penalty phase. The sum total of the mitigating evidence that Brogna offered at sentencing required less than three pages of a double-spaced transcript and relates only to Lam-bright’s behavior in jail. Although there were signs at the time that Lambright suffered from mental illness, his counsel failed to present any mitigating psychiatric testimony. Instead, in contravention of his obligation to his client, he relied solely on the probation officer’s report and that of the court-appointed psychologist to present any mitigating evidence regarding Lambright’s mental condition, history of drug dependency, and disadvantaged childhood. Cf. Correll, 465 F.3d at 1012. Brogna failed to call any of Lambright’s family members or friends to testify about his abusive childhood, his history of mental instability, his suicide attempts, or his drug use. Nor was any testimony offered to humanize Lambright, or to situate the crime within Lambright’s troubled history. Instead, Brogna promised the court that he had “one witness who will be very brief.” This sole witness, a guard at the Pima County Jail, testified that while Lambright was incarcerated the guard had experienced no “problems with him” and that Lambright had “been respectful, courteous, and cooperative.” Aside from this brief testimony, no mitigating evidence was introduced by Brogna. Instead, he submitted a two-and-a-half-page memorandum listing various potentially mitigating circumstances, including (1) the fact that a deal was cut with Lambright’s accomplice, Kathy Foreman, (2) Lambright’s “lack of a prior violent criminal record,” (3) “the Defendant’s record while incarcerated,” (4) that Lambright “is amenable to some type of treatment,” (5) the fact that his “military record is a good one,” and (6) that “[t]here is no record indicating he was a ‘juvenile delinquent.’ ” However, none of the potentially mitigating circumstances discussed in the memorandum was described in any detail, nor did Brogna point to any evidence in the record or provide any factual support for his assertions. Cf Correll, 465 F.3d at 1013(noting that no evidence was offered to support the assertions made in trial counsel’s brief sentencing memorandum).
Counsel’s argument at the close of the sentencing proceeding was similarly brief and uncompelling. Brogna merely asked the court to consider the fact that Kathy Foreman, a willing participant in the crime, received no punishment, that “[t]here isn’t anything in Mr. Lambright’s background that establishes the fact that he is a sociopathic personality,” and that Lambright did not have a violent history. He did not even mention Lambright’s unfortunate childhood, his history of substance abuse, or the mental problems that led him to two suicide attempts, a bizarre episode in which he wielded a butcher knife outside his sister’s door, as well as psychiatric hospitalization. Brogna did not ask for leniency or mercy; instead, he merely requested, without any elucidation, *1120that the court impose a sentence of life in prison. Accordingly, Brogna’s presentation at the penalty phase of Lambright’s trial, like his mitigation investigation, fell far below the level of representation to which Lambright was entitled under the Sixth Amendment. Cf. Correll, 465 F.3d at 1013; Stankewitz, 365 F.3d at 716-17(finding ineffective assistance where counsel offered the very general and un-compelling testimony of six witnesses, and merely summarized briefly petitioner’s life history “focus[ing] little on the actual details of [his] life” during closing argument).
The state argues, and the district court held, that Brogna’s performance was not deficient because his brief memorandum, along with the pre-sentence report and the report of the court’s mental health expert, Dr. Hinton, informed the sentencing judge that Lambright experienced a “bleak” childhood, that he had a history of drug abuse, and that he experienced mental health problems. Specifically, the district court held that Brogna did not perform deficiently because he
referenced, albeit in abbreviated form, many of the same mitigating factors identified in these proceedings ... Other factors [not mentioned by Brogna,] such as [cursory references to] the suicide attempts, hospitalization and abuse of drugsf,] were already before the sentencing judge and were considered in the weighing process. The Court therefore concludes that Brogna’s presentation was not constitutionally deficient.
The district court’s holding contravenes long-established Ninth Circuit law. As this court has often made clear, counsel’s duty to investigate all potentially mitigating evidence related to a defendant’s mental health, family background, and prior drug use and to provide the sentencing court with a full presentation of the evidence that might lead the sentencer to spare his client’s life is not discharged merely by conducting a limited investigation of these issues or by providing the sentencing court with a cursory or “abbreviated” presentation of potentially mitigating factors. Stankewitz, 365 F.3d at 716(“[C]ounsel’s duty to conduct a thorough investigation ... is not discharged merely by presenting some limited evidence.”); see also Wiggins, 539 U.S. at 524, 123 S.Ct. 2527 (finding ineffective assistance and prejudice where “counsel abandoned their investigation of petitioner’s background after having acquired only rudimentary knowledge of his history from a narrow set of sources”); Correll, 465 F.3d at 1015 n. 5 (holding that although “the bare facts of [petitionersj’s troubled past were ... presented to the court, without further investigation and presentation of contextual evidence and argument, such facts served only to demonize [petitioner] rather than to mitigate the appropriateness of imposing the death penalty for his actions.”); Douglas, 316 F.3d at 1090 (finding ineffective assistance where trial counsel “introduced some of [the defendant’s] social history, [but] did so in a cursory manner that was not particularly useful or compelling.”). To the contrary, “[w]hen it comes to the penalty phase of a capital trial, ‘[i]t is imperative that all relevant mitigation information be unearthed for consideration.’ ” Douglas, 316 F.3d at 1088 (quoting Caro v. Calderon, 165 F.3d at 1227). Only after a thorough investigation can a less than complete presentation of mitigating evidence ever be deemed reasonable, and only to the extent that a reasonable strategy supports such a presentation. Wiggins, 539 U.S. at 527-28, 123 S.Ct. 2527; Stankewitz, 365 F.3d at 719. Counsel may not rely for the development and presentation of mitigating evidence on the probation officer and a court appointed psychologist. See Correll, 465 F.3d at 1012 (finding “critical error” in counsel’s decision to allow the trial judge to make a *1121sentencing determination based on the state’s evidence and the pre-sentence report). The responsibility to afford effective representation is not delegable to parties who have no obligation to protect or further the interests of the defendant.
In sum, the record makes plain that Brogna’s limited, cursory, and incomplete presentation of mitigating evidence did not take place after “ ‘all relevant mitigation information [was] unearthed for consideration,’ ” Douglas, 316 F.3d at 1088 (quoting Caro v. Calderon, 165 F.3d at 1227). Instead, Brogna spent only five and a half hours, following Lambright’s conviction, in preparation for the penalty phase (including three hours talking to the probation officer, preparing his subpoena for the jail guard, and dictating the memo to the court). Meanwhile, he ignored “tantalizing indications in the record ... that ‘would [have] le[]d a reasonable attorney to investigate further.’ ” Stankewitz, 365 F.3d at 720(quoting Wiggins, 539 U.S. at 527, 123 S.Ct. 2527). Compounding his failure to investigate, Brogna then failed to draw the court’s attention to the scant mitigating evidence that was before the court through the pre-sentence report and Dr. Hinton’s report, or to argue that this evidence should be given mitigating weight. Such representation falls far below that which any reasonably competent attorney would provide in a capital case. Accordingly, we hold that the district court erred in concluding that Brogna’s performance at the penalty phase was not deficient.
B. Prejudice
To establish prejudice, Lambright must demonstrate that there is “a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Strickland, 466 U.S. at 694, 104 S.Ct. 2052. A reasonable probability is one “ ‘sufficient to undermine confidence in the outcome,’ ” but is “less than the preponderance more-likely-than-not standard.” Summerlin, 427 F.3d at 640, 643(quoting and citing Strickland, 466 U.S. at 693-94, 104 S.Ct. 2052). Accordingly, “[i]n establishing prejudice under Strickland, it is not necessary for the habeas petitioner to demonstrate that the newly presented mitigation evidence would necessarily overcome the aggravating circumstances.” Correll, 465 F.3d at 1018(citing Williams v. Taylor, 529 U.S. 362, 398, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000)); see also Rompilla v. Beard, 545 U.S. 374, 393, 125 S.Ct. 2456, 162 L.Ed.2d 360 (2005) (“[A]lthough we suppose it is possible that [the sentencer] could have heard it all and still decided on the death penalty, that is not the test.”). Instead, in evaluating prejudice, we must “compare the evidence that actually was presented to the [sentencer] with the evidence that might have been presented had counsel acted differently,” Bonin v. Calderon, 59 F.3d 815, 834(9th Cir.1995), and evaluate whether the difference between what was presented and what could have been presented is sufficient to “undermine confidence in the outcome” of the proceeding, Strickland, 466 U.S. at 694, 104 S.Ct. 2052.
Applying these standards, both this court and the Supreme Court have consistently held that counsel’s failure to present readily available evidence of childhood abuse, mental illness, and drug addiction is sufficient to undermine confidence in the result of a sentencing proceeding, and thereby to render counsel’s performance prejudicial. See, e.g., Rompilla, 545 U.S. at 390-91, 125 S.Ct. 2456; Ainsworth v. Woodford, 268 F.3d 868, 878 (9th Cir.2001); Smith v. Stewart, 140 F.3d at 1271.
In Smith v. Stewart, we held that prejudice had been established where defense counsel did very little investigating, presented no mitigating evidence, and made only “a few asthenic comments” on his *1122client’s behalf in spite of the fact that additional mitigating evidence “was rather near the surface.” Smith v. Stewart, 140 F.3d at 1271. Smith was convicted of shooting a convenience store clerk while in the process of robbing the store, and the state sought the death penalty. Id. at 1267-68. In preparation for sentencing, Smith’s counsel spoke with Smith and Smith’s mother. Id. at 1269. He asked them “about [Smith’s] growing up years, but did not discover any difficulties worth mentioning, and does not recall having been made aware of any treatment Smith might have received at a medical facility,” most likely because he “asked nothing more than a few generalized questions and conducted none of the real probing for information that legal praxis assumes and even demands.” Id. In light of the evidence Smith presented in support of his habeas petition, we noted that “counsel could have, at least, pointed to Smith’s sociopathic personality, his bad drug history, his change in personality after a large drug overdose, and his fine set of family relationships at the time” as evidence in mitigation. Id.
In assessing prejudice, the court noted that the fact that Smith had a sociopathic personality is mitigating, given that “[t]he Arizona Supreme Court has made it clear that an antisocial personality disorder (so-ciopathic disorder) is a mitigating factor, even if it does not come up to the level of a factor specifically listed in the Arizona sentencing statute.” Id. at 1270 (citing State v. Thornton, 187 Ariz. 325, 929 P.2d 676, 685-86 (1996) and State v. Stokley, 182 Ariz. 505, 898 P.2d 454, 470-71, 473 (1995)).
Similarly, Smith’s long use of drugs is a factor that could have had a mitigating effect.... Along the same line, a major personality change can amount to a mitigating factor in Arizona. Smith has presented some evidence that a change of that nature came over him after he had ingested a large amount of POP in 1980.... Finally, Smith points to his good family relationships. That was before the sentencing judge through the presentence report, but in a very mild form. With a little effort it could have been developed through evidence or argument, and could have put Smith himself in a somewhat different light.
Id. at 1271, 898 P.2d 454(citing State v. Gallegos, 178 Ariz. 1, 870 P.2d 1097, 1113 (1994); State v. Rockwell, 161 Ariz. 5, 775 P.2d 1069, 1079-80 (1989)). Assessing this evidence, we noted that
these factors, even as now developed, do not demand a life sentence rather than a sentence of death. On the other hand, while the facts of this case are bad enough to disturb even a jaded observer, they do not reach the level of those in cases where the aggravating facts were so overwhelmingly horrifying that it was highly improbable that mitigating factors of any ordinary stripe would help.... In fine, and with all due respect to those who have heard this case before us, our confidence in the sentencing result has been undermined.
Id. (citations omitted).
Here, as in Smith, significant mitigating evidence was available but was never developed or presented to the sentencing judge. Here, too, there was some limited and unsubstantiated mitigating evidence before the court through the presentence report and Dr. Hinton’s report, but counsel failed to develop it or even argue that the court should consider it. Indeed, the evidence that could have been presented or relied upon at Lambright’s sentencing hearing is strikingly similar to the evidence the absence of which undermined our confidence in the outcome of the sentencing proceeding at issue in Smith. Specifically, the failure to offer the following evidence that could and should have *1123been developed and presented was prejudicial.
1. Lambright’s Traumatic Childhood
The district court found that facts related to Lambright’s childhood “w[ere] largely identified to the sentencing judge,” and based on that finding concluded that Brog-na’s failure to develop this evidence was not prejudicial. This finding is clearly erroneous, that the resulting conclusion lacks support. While the pre-sentence report and Dr. Hinton’s report did provide some very limited information about Lam-bright’s upbringing, the picture of his childhood presented in the reports was inaccurate, undeveloped, and unsubstantiated. Had Brogna conducted a reasonable investigation, he would have been able to present compelling evidence showing that Lambright was subjected to near constant emotional and physical abuse by his mother throughout his childhood, and that his father did nothing to stop this abuse. Specifically, testimony could have been offered to show that his mother frequently hit, kicked, or whipped him, and that these beatings were a daily occurrence during extended periods of his childhood. Although the sentencing court may have been aware that Lambright’s mother was “very strict” and was “ill-equipped with the necessary parenting skills” to raise two children, given that these facts were briefly mentioned in the pre-sentence report, no evidence regarding the significant physical abuse Lambright experienced as a child was developed or presented to the sentencing court.
Similarly, although the pre-sentence report noted that Lambright’s mother was hypochondriacal and strictly Pentecostal, it did not convey the fact that she spent the majority of Lambright’s childhood in bed claiming to be afflicted by various illnesses, nor that she was profoundly addicted to prescription drugs. The sentencing judge was also unaware of the fact that Lam-bright’s mother forced him to take Valium and sleeping pills when he acted up or had too much energy.
Although the pre-sentence report and Dr. Hinton’s report noted that Lam-bright’s family moved frequently during his childhood, these reports did not effectively convey the extent to which these moves prevented Lambright from developing stable social relationships as a young child — relationships that might have mitigated the effect of his abusive home environment. The court was also unaware that Lambright never attended any school for more than a year and, as a result, had difficulty forming friendships with other children and was a frequent target of schoolyard bullies.
The sentencing judge was likewise unaware of the poverty that Lambright experienced while growing up. In the presen-tence report, the probation officer noted that according to Lambright and his older sister, they were raised in a lower-middle class family. The sentencing judge did not know that their constant moves were necessitated by their father’s struggle to maintain steady employment, and that the family once was forced to move into a rat-infested house in which the walls and ceilings were lined with cardboard to block holes. He likewise did not know that they often lived in houses with no running water or indoor plumbing.
Accordingly, Lambright has demonstrated that compelling mitigating evidence could and should have been presented regarding his childhood, evidence that, at least in combination with the evidence that Brogna failed to present regarding Lam-bright’s long-term drug abuse and impaired mental condition, undermines our confidence in the result. Such failure was, accordingly, prejudicial.
*11242. Lambright’s Long-Term, Drug Use and Dependency
The district court concluded that “the sentencing judge was already well aware of [Lambright’s history of drug abuse] through the pre-sentence report. There was not much more evidence that Brogna could have uncovered that would have been helpful or that the sentencing judge did not already know with respect to [Lambright’s] drug use.” This conclusion is likewise clearly erroneous in light of the evidence presented at the evidentiary hearing. Although the pre-sentence report and Dr. Hinton’s report did mention Lambright’s drug use, together the reports contained only a few sentences on this point, and provided incomplete, un-compelling, and uncorroborated information. Neither report contained any information about the extent to which these drugs affected Lambright’s development or his social relationships, nor the nature of their effect on him. Had Brogna properly developed evidence regarding Lam-bright’s long history of drug dependency, he would have been able to show that drug and alcohol abuse were rampant in Lam-bright’s family and that, as noted above, it was his mother who first exposed him to drug abuse by forcing him to take sedatives when he was a child. Brogna could have presented further evidence regarding the fact that there were periods in Lam-bright’s life during which he took large quantities of methamphetamine, staying awake for weeks at a time. He could have presented evidence about the negative impact drug abuse had on his ability to form relationships, and could have provided expert testimony about the long-term psychological impact on him of his extreme methamphetamine abuse. At the eviden-tiary hearing, even the state’s expert, Dr. Lang, diagnosed Lambright with profound substance dependency.
Accordingly, significant evidence related to Lambright’s long-term drug addiction and its effect upon him was available but was not presented to or urged upon the sentencing court, evidence which, in combination with the evidence Brogna failed to introduce regarding Lambright’s childhood and his mental health problems, undermines our confidence in the result. The failure to present such evidence was, accordingly, prejudicial.
3. Lambright’s Mental Health Problems
The district court found that Lam-bright’s “testimony with respect to the trauma underlying his expert’s PTSD diagnosis (the alleged combat experience) is less than credible.” On the basis of the testimony and other evidence offered at the evidentiary hearing, we conclude that this finding is not clearly erroneous.4 However, because it considered only the question whether Lambright had proved that he suffered from PTSD, the district court failed to discuss evidence establishing that Lambright did suffer from various other mental health problems. Moreover, irrespective of whether Lambright suffered a traumatic combat experience in Vietnam, there was testimony presented at the evidentiary hearing that he was a very different person once he returned from his tour of duty. Cf. Smith v. Stewart, 140 F.3d at 1271(“[A] major personality change can amount to a mitigating factor in Arizona.”).
*1125In addition, testimony at the evidentiary hearing described Lambright’s lifelong struggle with depression. Dr. Morenz, Lambright’s mental health expert, testified that Lambright has experienced varying levels of depression for most of his life, and at times has suffered from “major depression.” Even Dr. Lang, the state’s expert, diagnosed him with a depressive disorder. As a result of his profound depression, Lambright twice attempted suicide. Moreover, on at least one occasion, Lambright experienced hallucinations and was hospitalized as a result.
Expert testimony presented at the evi-dentiary hearing also suggested that Lam-bright may suffer from a personality disorder. Dr. Lang, the state’s expert, whose assessment of Lambright’s mental health problems the district court found credible, concluded that Lambright suffers from a personality disorder not otherwise specified with antisocial, borderline, and inadequate features. This diagnosis, if properly developed and explained to the sentencer, would have had a mitigating effect under Arizona law, as noted in Smith v. Stewart, 140 F.3d at 1270(citing Thornton, 929 P.2d at 685-86; Stokley, 898 P.2d at 470-71, 473).
While the sentencing court was provided with some limited information about Lam-bright’s mental health problems through the pre-sentence report and Dr. Hinton’s report, the information contained in those reports was conclusory and based solely on Lambright’s own accounts of his history. No medical records were offered to substantiate his reports of suicide attempts or psychiatric hospitalization, and no mental health professional was called to testify about the extent and nature of his mental health problems. Thus, while certain basic information about Lambright’s mental health problems was presented to the sentencing judge, this information was incomplete, not eompellingly presented, and buried in otherwise damning reports.
Moreover, because Brogna apparently failed to recognize that Dr. Hinton’s diagnosis of antisocial personality disorder is a mitigating factor under Arizona law, see Smith v. Stewart, 140 F.3d at 1270, he never urged the sentencing judge to give this diagnosis mitigating weight. Indeed, Brogna never argued that the scant mental health evidence that was before the court should be considered in mitigation. Instead, he suggested that Lambright did not suffer from the mental health problem diagnosed by Dr. Hinton.
Brogna’s failure to identify the minimal mitigating evidence in the record and to urge it in mitigation likewise applies to Lambright’s disturbed childhood and his history of drug abuse. Brogna never pointed out to the sentencing judge that either of these conditions constituted mitigating factors, never urged that the judge treat them as such, and never explained the weight that should be given them.
We do not underestimate the importance of the role of counsel in the adversarial process. The sentencing judge cannot be expected to comb the record looking for mitigating factors, particularly where the minimal evidence that exists is buried in reports that are on the whole strongly unfavorable to the defendant. When a defense lawyer fails to identify crucial mitigating factors and fails to explain why those factors should affect the judge’s decision, that failure alone may, in some instances, be prejudicial. To the extent that the state argues that the minimal mitigating information that was before the court principally as a result of the submission of the pre-sentence report and Dr. Hinton’s report defeats Lambright’s claim of prejudice, we rejected that argument in Correll. There, we held that
*1126[w]hile the bare facts of Correll’s troubled past were indeed presented to the court [through the pre-sentence report and defense counsel’s brief sentencing memorandum], without further investigation and presentation of contextual evidence and argument, such facts served only to demonize Correll rather than to mitigate the appropriateness of the death penalty for his actions.
Correll, 465 F.3d at 1015 n. 5.
In sum, there is a significant quantity of “classic” mitigating evidence that was nevr er presented or urged upon the sentencing judge on account of Brogna’s failure to adequately investigate and prepare Lam-bright’s sentencing defense. Had counsel conducted a reasonable investigation with respect to Lambright’s childhood, mental health, and history of drug abuse, had he effectively presented the information he would have uncovered, and had he then made the appropriate arguments to the sentencing judge, there is a reasonable probability that the result of the sentencing proceeding would have been different; put plainly, Brogna’s failure to provide effective assistance of counsel undermines our confidence in the result.
The prejudicial nature of Brogna’s failure to develop and adequately present available mitigating evidence becomes particularly “apparent when we consider the effect of the error under Arizona law. At the time of the penalty phase proceedings, Arizona law mandated the death penalty [where one or more aggravating factors were present] ... if there was no mitigating evidence.... Thus, [a lawyer’s] failure to present any evidence in mitigation [would] ‘all but assure[ ] the imposition of a death sentence under Arizona law.’ ” Id. at 1012 (quoting Summerlin, 427 F.3d at 640); see also Smith v. Stewart, 140 F.3d at 1268, 1270(stating that, in light of Arizona’s statute, counsel’s “few asthenic comments” at sentencing amounted to a “virtual admission that the death penalty should be imposed upon his client”).
The effect of Arizona’s death penalty statute is particularly striking in the case at hand given the sentencing judge’s statements at the hearing in which he announced his penalty verdict. In sentencing Lambright to death, the sentencing judge stated that
the very factors which make murder a terrible crime make the death penalty a terrible punishment.
THE COURT has been asked if “justice” requires the maximum penalty; the question is answered by quoting the statute!:] “... The Court ... shall impose a sentence of death if it finds one or more aggravating circumstances and finds no mitigating circumstances sufficiently substantial to call for leniency.”
THE COURT HAS SO FOUND, and the LAW demands the maximum penalty ...
(emphasis in original). Thus, the sentencing judge’s own statements make clear that, given Brogna’s failure to present any significant mitigating evidence, the judge felt legally compelled to impose the death penalty in this case. Had Brogna presented the mitigating evidence that would have been unearthed through a reasonable investigation of the mitigating circumstances, there is a reasonable probability that the result of the sentencing judge’s penalty deliberations would have been different.
Finally, this is not a case in which there were a multitude of aggravating factors that Lambright would have had to overcome in order to receive a life sentence. To the contrary, the sole aggravating factor found by the sentencing judge was the fact that the crime here at issue was “committed in an especially heinous, cruel, or depraved manner.” As we noted in our *1127opinion remanding this case for an eviden-tiary hearing,
[w]e have previously found that the prejudice requirement is met where “defense counsel effectively presented no mitigating evidence at sentencing, despite the presence of aggravating factors.” ... [I]n this case the State argued that only one aggravating factor existed. Although the offense in this case certainly was brutal and sadistic, the Supreme Court in Williams recently noted that “[mjitigating evidence unrelated to dangerousness may alter the jury’s selection of penalty, even if it does not undermine or rebut the prosecution’s death-eligibility case.” ... Evidence of mental disabilities or a tragic childhood can affect a sentencing determination even in the most savage case.
Lambright III, 241 F.3d at 1208 (citations omitted).
We have held that, in light of the Supreme Court’s decisions in Wiggins and Williams v. Taylor, it is clear that “the presentation of mitigating evidence is vital even where ... the aggravating evidence is powerful.” Stankewitz, 365 F.3d at 714. In Ainsworth, for example, where we held that counsel’s failure to present evidence very similar to that which should have been presented in Lambright’s case was prejudicial, the offense was at least as, if not more, vile than the offense at issue here. Ainsworth, 268 F.3d at 870-71. Ainsworth approached the victim as she sat in her parked car, and shot her in the left hip with a .45 caliber handgun. Id. at 870. The bullet passed through her pelvis and lodged against her right hip. Id. During the next twenty-four hours, Ainsworth confined the suffering victim within her car, ignored her pleas for help and, at one point, put her in the trunk because he was tired of hearing her moan and cry. Id. Then, after removing her from the trunk, he raped her. Id. Twenty-four hours after she was shot, the victim finally died and Ainsworth dumped her body and abandoned her car. Id. at 871. In spite of the savageness of this offense, and in spite of other significant aggravating factors present in that case, we held that Ainsworth’s counsel’s failure to introduce available mitigating evidence regarding “Ainsworth’s troubled childhood, history of substance abuse, and mental and emotional problems” was prejudicial. Id. at 875, 878.
Indeed, we have held consistently that even in cases involving particularly heinous murders, a defendant can be prejudiced by an attorney’s failure to investigate and present mitigating evidence. See Hovey v. Ayers, 458 F.3d 892, 930 (9th Cir.2006) (noting that even where a defendant has been convicted of horrible crimes, a sen-tencer may still choose to impose a life sentence when confronted with mitigating evidence); Douglas, 316 F.3d at 1091 (finding that “[t]he gruesome nature of the killing did not necessarily mean that the death penalty was unavoidable.”); Smith v. Stewart, 189 F.3d at 1013(“[T]he horrific nature of the crimes involved here does not cause us to find an absence of prejudice.”); Hendricks v. Calderon, 70 F.3d 1032, 1044 (9th Cir.1995) (concluding that “despite ... substantial evidence of aggravation, ... the failure ... to present mitigating evidence rendered the sentencing hearing neither fair nor reliable”).
In sum, Brogna’s failure to conduct a thorough investigation of all potentially mitigating evidence was unreasonable, especially in light of the information available to him, and fell far below standards of professional conduct. Because the evidence counsel failed to uncover and present to the sentencing judge was highly significant, because there was but one factor in aggravation in this case, and because Brogna’s failures were particularly prejudicial in light of Arizona law, Lambright has adequately “undermined [our] confi*1128dence in the outcome” of the penalty phase. See Strickland, 466 U.S. at 694, 104 S.Ct. 2052. For these reasons, we hold that the district court erred in finding that Lambright was not prejudiced by Brogna’s ineffective representation.
CONCLUSION
The district court erred in concluding that Lambright did not receive ineffective assistance of counsel at the penalty phase of his trial and further erred in holding that, even if he did, he suffered no prejudice as a result. The sentence imposed is vacated. Further sentencing by the state court shall be conducted in conformance with applicable law.
REVERSED and REMANDED for issuance of a writ of habeas corpus.
. A fuller description of these events appears in our earlier en banc opinion. Lambright v. Stewart, 191 F.3d 1181 (9th Cir.1999).
. We need not reach Lambright’s argument that the district court erred in drawing adverse inferences based on his refusal to answer questions related to the crime as any error in drawing these inferences was harmless. Neither of the two inferences drawn is of consequence to our analysis of the ineffective assistance of counsel claim. The first adverse inference — that Lambright was not on drugs when he committed the offense — is irrelevant given that Lambright does not argue on appeal that he was in fact on drugs at that time but only that his history of drug addiction is a mitigating factor in and of itself. The second inference — that Lam-bright’s alleged Vietnam experience did not affect his behavior at the time of the offense— is likewise irrelevant given that we hold, for purposes of this appeal, that the district court did not clearly err in finding that the allegedly traumatic experience never occurred.
. At one point in its order, the district court concludes that "Brogna interviewed Petitioner's only sibling extensively!.]” The record does not support the court’s finding that the interview was "extensive." Brogna testified that his recollection of his sole interview with Lambright's sister was "very vague! ]," that the interview focused on Lambright's life shortly before and after the crime, and that he doesn't recall whether he talked with her about Lambright's upbringing, social history, or mental health problems. Lambright’s sister’s testimony about this interview is inconsistent, most likely due to the fact that she suffers from memory problems as the result of a stroke. Thus, we cannot accept the district court's finding that this interview was "exten*1119sive.” In any event, this finding does not affect our decision. Even if counsel had spoken to Lambright's sister about his childhood, history of drug abuse, and mental health problems prior to the guilt phase of Lam-bright's trial, his total failure to do any other investigating or to follow up on what he might have learned from her, as well as his failure to present evidence of any of these mitigating circumstances to the sentencing judge, would render his assistance ineffective.
. The district court erred, however, in declaring that even if it were “persuaded that the Petitioner suffered from PTSD, Petitioner has failed to show how the disorder affected his criminal conduct.” As we explained supra at Part II, there need not be a causal connection between a factor, such as mental illness, disturbed childhood, or drug dependency, and the criminal offense in order for it to be mitigating.