dissenting:
The majority errs in affirming the death sentence of Cal Brown. Brown’s attorneys made a highly deficient presentation regarding the most important aspect of his mitigation case — that Brown suffered from a serious mood disorder that was treatable with lithium. Had they performed at a level consistent with the prevailing professional norms, the jury would have been aware that with the proper medication, Brown would have had greater control over his impulses and would have been less likely to commit the crime. The jurors *1038would also have learned that with the proper treatment, Brown would not be a threat in the future. I cannot say with confidence that had the jury possessed this information it would have voted unanimously to impose the death penalty.
Despite Brown’s counsel’s knowledge that the medical treatment of their client’s mental disorder would be critical to their mitigation theory and despite the fact that counsel were told by their own expert psychologist, Dr. Maiuro, that they should retain a psychiatrist to testify on that subject, they failed to do so. Instead, Dr. Maiuro remained their sole expert, and he was forced to admit on the stand that he was not qualified to prescribe medication and could not testify about whether Brown’s disorder should be treated with lithium or whether proper treatment might have prevented the crime. Additionally, Brown’s attorneys failed to call his counselor Sally Schick, the only mental health professional who had the opportunity to observe and treat him for a sustained period of time, and the one who first recommended that he receive lithium treatment. Finally, Brown’s attorneys failed to cross-examine Dr. Brinkley, the State’s expert psychiatrist, after his devastating testimony that Brown did not have any disorder that could be treated with lithium; by implication, Dr. Brinkley suggested that Brown did not have a serious mental disorder and that no medical treatment of any kind could have had a positive effect on his conduct. Moreover, because of counsel’s failure to cross-examine Dr. Brinkley, the jury did not learn that his testimony was based on his examination of Brown’s prison records alone. The prosecutor took advantage of Brown’s counsel’s ineptitude and seriously undermined the case for mitigation by emphasizing in his closing argument the lack of qualifications of Brown’s expert and reminding the jury that Dr. Brinkley’s testimony had not been challenged.
Brown’s counsel’s deficient performance prejudiced him because the jury never heard the most convincing argument for why their client’s life should be spared. Brown’s mitigation case consisted of two related contentions. The first was that he had a mental disorder. If defense counsel had put on a proper case to try to persuade the jury that Brown suffered from a serious mental illness, at least some of the jurors might have understood his crimes to be a manifestation of that disorder and perceived him to have had less control over his heinous acts than a mentally healthy person.
The second part of the mitigation theory was that Brown’s disorder could have been effectively managed medically with lithium. The lithium issue is important because it reinforces the mental illness argument and because it illustrates the extent to which Brown’s actions were influenced by that disorder. Lithium is a strong medication that is closely linked to serious mental illness in the popular consciousness. Any disorder that required lithium treatment might have been considered quite severe in the eyes of the jurors. Thus, they could more readily have understood that the defendant’s conduct was influenced by a mental disorder that could be treated medically. If the jurors had heard the testimony of a psychiatrist like Dr. Scher, who would have said that lithium would have made Brown less likely to commit the crime, at least some of them might have attributed less of the blame for the crime to Brown’s character and more of it to his physiological condition. Additionally, although the issue of future dangerousness was never argued at trial, the jury likely would have taken into account that if Brown were properly treated, he would no longer pose as great a threat to society. Counsel’s deficient performance prejudiced Brown because it rendered both prongs of the mitigation theory unconvincing. There *1039is no question that defense counsel failed to prove the second part of the mitigation case — that Brown’s illness could be effectively treated with lithium and that such treatment would have made Brown less likely to commit crimes. The defense’s only expert witness admitted that he was unqualified to testify about lithium. The State emphasized this in cross-examination and in its closing argument and its expert testified that Brown did not have a disorder that could be treated with lithium. The jury had no choice but to conclude that lithium would have had no effect on Brown, despite counsel’s claim to the contrary in his opening statement.
Defense counsel’s deficient performance also undermined the first claim — that Brown had a serious disorder. Brown’s lawyers failed to cross-examine Dr. Brinkley and thus allowed the prosecutor to argue to the jury that the defense must agree with Dr. Brinkley’s assessment that Brown did not suffer from any disorder that could be treated with lithium. Without cross-examination, the jury was not informed of the fact that Dr. Brinkley had never met with Brown, as is customary before diagnosing mental disorders, and that he did not review all of Brown’s records including those that contained information that he later admitted could have been important to his diagnosis. All it heard was that Dr. Brinkley concluded that Brown did not have a mental illness that would warrant treatment with lithium and that, as the prosecution argued, the defense did not challenge that conclusion. The jury was also told by the prosecutor, with considerable effectiveness, that Dr. Maiuro was not qualified to testify regarding lithium. The jury might well have concluded that if Dr. Maiuro was not as qualified as Dr. Brinkley to testify about lithium, he was not as qualified to testify about mental disorders in general. Given Dr. Brinkley’s unimpeached testimony and the lengthy and contentious cross-examination of Dr. Maiuro, the jury likely credited Dr. Brinkley’s testimony over Dr. Maiuro’s and may have concluded that Brown did not suffer from a mental disorder at all. In reality, the absence of a defense psychiatrist and the failure to cross-examine Dr. Brinkley resulted in the almost total undermining of defendant’s mental illness mitigation argument. Even if the jury believed that Brown had psychological problems, they likely assumed that his disorders could not have been particularly severe if they did not require any lithium treatment. (No one suggested that any other form of medical treatment would have been appropriate.)
There is a reasonable probability that if defense counsel had presented the testimony of a psychiatrist, as well as that of Brown’s counselor, and if they had cross-examined Dr. Brinkley, they would have raised a sufficient question in the mind of at least one juror as to whether Brown suffered from a serious mental disorder that could have been controlled with medical treatment, and that at least one juror would have concluded that the prosecution had not proven beyond a reasonable doubt that there were no sufficient mitigating circumstances to warrant the imposition of a life sentence. Counsel’s failures prejudiced Brown by casting substantial doubt on the most important aspect of his mitigation case. Such a deficient and prejudicial performance severely undermines one’s confidence in the outcome of the penalty proceeding. I would reverse Brown’s death sentence so that the state could retry the penalty phase or impose a sentence of life without parole.
I. Failure to Call a Psychiatrist
Brown’s defense counsel knew that the lithium issue would be an important part of their case, and their expert witness, Dr. Maiuro, a psychologist, advised them that *1040he was not qualified to testify about lithium. He recommended that they retain a psychiatrist who was. Nevertheless, Brown’s lawyers not only failed to obtain the services of a psychiatrist to testify as an expert at the penalty phase, they never even consulted one. This failure was objectively unreasonable under then-applicable prevailing professional norms.
According to the ABA Guidelines at the time of trial, in preparation for the penalty phase, counsel should consider, “expert witnesses to provide medical, psychological, sociological or other explanations for the offense(s) for which the client is being sentenced, to give a favorable opinion as to the client’s capacity for rehabilitation, etc. and/or to rebut expert testimony presented by the prosecutor.”1 Defense counsel’s primary mitigation argument at the penalty phase was that Brown suffered from a mood disorder for which he did not receive adequate medical treatment — treatment that could have lessened his symptoms and improved his impulse control. The testimony of a psychiatrist was necessary to show that Brown’s disorder could have been effectively managed medically and to rebut the testimony of Dr. Brinkley that Brown did not suffer from a mood disorder for which lithium was appropriate.
Counsel was on notice that a psychiatrist would be needed to testify about the proper medical treatment of Brown’s disorder. Lin Marie Hupp, one of Brown’s attorneys, testified at the evidentiary hearing that defense counsel understood the importance of the lithium issues because “[hjaving something that is physiologically based is sometimes easier for jurors to understand ... that you can give somebody a pill for and fix them is sometimes easier for jurors to understand than just ... the person is a bad person.” Hupp’s testimony demonstrates that the lawyers understood that if they showed that Brown’s disorder could be medically treated, it would enable the jurors to comprehend the seriousness of the disorder, understand that it contributed to his conduct, and thus recognize its mitigating effect. Additionally, William Schipp, the staff social worker, explained to Brown’s attorneys the difference between psychologists and psychiatrists, so they understood that Dr. Maiuro, a psychologist, would be unable to testify about the proper medical treatment of Brown’s disorder. Moreover, Dr. Maiuro personally advised the defense that he was unable to testify about medical matters, recommended that they consult a psychiatrist, and gave them referrals. Perhaps the best evidence that defense counsel understood that it was crucial for them to hire an expert psychiatrist is the fact that they attempted to do so when, after receiving Dr. Maiuro’s advice, they contacted Dr. Brinkley and tried to enlist his services.
At the evidentiary hearing, none of the defense attorneys offered a reason for their failure to seek a psychiatrist either months before trial or after Dr. Brinkley informed them that he was unavailable because he was testifying for the State. Kern deven testified that he and his colleagues knew about Brown’s mood disorder six to eight months before trial and that “we were aware that we would like to have another kind of an expert ... a psychiatrist.” When Cleven was asked why they did not contact one, he said: “I can’t for the life of me think of why it was that late in the game and we didn’t have a psychiatrist on board yet. I wish I could provide you with a reason why that was, I just can’t.” He also recalled that there was no impediment to contacting another *1041psychiatrist to testify once they knew that Dr. Brinkley was unavailable. Terry Lee Mulligan testified that he did not recall there being “a specific decision made” or “conversations about” consulting any other psychiatrists.
Given the overwhelming evidence that defense counsel knew that it was necessary to consult a psychiatrist and present his expert testimony in order to establish that Brown had a disorder that could and should have been treated with lithium, and given that counsel can offer no reason why they failed to consult and present the testimony of a psychiatrist, I would hold that counsel’s failure to retain such an expert was unreasonable under existing professional norms at the time and thus amounted to deficient performance under Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984).
The majority suggests strategic justifications for defense counsel’s decision not to present the testimony of a psychiatrist at the penalty phase. Although we are highly deferential to counsel’s strategic decisions at trial, Strickland, 466 U.S. at 689, 104 S.Ct. 2052, counsel did not advance at the evidentiary hearing any of the explanations the majority gives for their failures. When counsel is unable to provide us with strategic reasons for trial decisions, we are not permitted to engage in “post hoc rationalization.” Wiggins v. Smith, 539 U.S. 510, 526-27, 123 S.Ct. 2527, 156 L.Ed.2d 471 (2003). Instead, we must evaluate counsel’s performance by the standard established in Strickland — “reasonableness under prevailing professional norms.” Strickland, 466 U.S. at 688, 104 S.Ct. 2052. Because there is no reasonable explanation for counsel’s failure to consult and retain a psychiatrist, the majority’s attempted rationalizations are unavailing.
The majority claims that defense counsel decided that they would hire a psychologist instead of a psychiatrist because “a psychiatric evaluation wouldn’t provide as much useful information as a psychological evaluation.” Maj. Op. at 1033. This is a puzzling conclusion given that defense counsel knew well in advance of trial that Brown’s medical treatment for his mental illness would be a critical mitigation issue and a psychologist would not be qualified to testify on that subject. Moreover, at the habeas corpus evidentiary hearing, Schipp testified that initially defense counsel decided to pursue a psychological evaluation for the reason cited by the majority. However, he offered no explanation as to why, after psychological testing showed that Brown had a mental disorder that could be treated with lithium, counsel did not consult a psychiatrist. At that point, all of the psychological tests in the world could not have produced the one “useful” piece of evidence that was necessary— expert testimony that Brown’s disorder could and should be treated with lithium. Additionally, the fact that defense counsel attempted to hire Dr. Brinkley belies the explanation that they did not believe a psychiatrist would be useful.
Nor, as the majority suggests, did Brown’s defense counsel decide not to seek an expert psychiatrist in order to avoid a continuance. Maj. Op. at 1034-35. None of Brown’s lawyers testified that seeking another psychiatrist would have required a continuance; indeed the first time this explanation was offered was in the State’s appellate brief — unsurprisingly with no citation to the record.2 Moreover, no one suggests any reason why seeking another psychiatrist would have required a continuance. Defense counsel sought to obtain *1042Dr. Brinkley’s services for the penalty phase on the first day of the guilt phase proceeding. Either they believed that they could hire Dr. Brinkley at that time to testify as an expert witness without creating a need for a continuance, or they believed that the testimony of an expert psychiatrist was more important than the timing of the penalty phase. If it was the latter, then failing to contact another psychiatrist after Dr. Brinkley advised them of his unavailability was obviously unreasonable. If it was the former, there is no reason why a different psychiatrist would have required more time than Dr. Brinkley. In that case, no continuance would have been necessary at all had they acted promptly. In any event, the psychiatrist whose services defense counsel obtained would not have had to testify until the penalty phase. Counsel could have contacted a different psychiatrist on the same day that Dr. Brinkley refused and that psychiatrist would have had just as much time to prepare his testimony as Dr. Brinkley. Nothing about Dr. Brinkley’s refusal of the request to serve as a defense witness raised any question as to the need for a psychiatrist or as to the merits of the mitigation argument. He simply advised counsel that he was unavailable because he had been hired by the prosecution. This should, if anything, have made it even more obvious that the defense needed a psychiatrist of its own.
Even had defense counsel believed that they would have been required to seek a continuance, this belief was not supported by any inquiry on their part. According to Schipp, defense counsel did not request him to contact any other psychiatrists to ask whether they could be ready in time. Mulligan also testified that his records indicated no further attempt to contact any psychiatrist after he learned that Dr. Brinkley was unavailable, although Dr. Maiuro testified that he probably recommended a few psychiatrists to defense counsel. Dr. Maiuro also stated that the defense did not contact him for additional referrals. The testimony shows that defense counsel abandoned their search for a psychiatrist after their first call to Dr. Brinkley failed, and that they did not inquire into whether another psychiatrist would be available or whether there would have been a need for a continuance. “[Sjtrategic choices made after less than complete investigation are reasonable precisely to the extent that reasonable professional judgments support the limitations on investigation.” Strickland, 466 U.S. at 691, 104 S.Ct. 2052.
Although the evidence overwhelmingly shows that defense counsel’s failure to hire a psychiatrist was not due to their desire to avoid a continuance, even if that had been their explanation, it would have been unreasonable. The majority argues that the defense wanted the jury to deliberate over the Christmas holiday and wanted to give the State less time for penalty phase preparation. Maj. Op. at 1034-35. As a preliminary matter, eleven’s testimony reveals that these were minor concerns at most. He testified that the Christmas holiday was “a silly point” and “not a major consideration for us.” He also admitted that “there was nothing remotely scientific” about the speculation that the jury might be more merciful over the Christmas holiday and there were no “focus groups or trial jurors or anything else” that supported any such thesis. If the Christmas holiday were a major consideration, which it was not, the defense’s armchair juror psychology would not be entitled to deference because it was not the result of proper investigation. Additionally, Cleven testified that he did not think that giving the State more time to prepare was “much of a consideration.”
The majority also claims that defense counsel did not call another psychiatrist *1043because they feared that he would render an unfavorable opinion and did not want to take the risk that they would be compelled to turn over a damaging report to the State. Maj. Op. at 1034-35. Not only is that argument unsupported by the record because no defense attorney testified that fear of an unfavorable report was the reason for their decision, but it is also nonsensical given that the lawyers actually attempted to hire Dr. Brinkley. The majority does not explain why counsel would try to hire Dr. Brinkley with no concern that he would render an unfavorable opinion, but, after he informed them that the State had hired him, would suddenly develop fears of a damaging diagnosis and elect not to contact another psychiatrist. Moreover, the record shows that defense counsel had a policy not to obtain a written report from their experts precisely in order to avoid such unwanted discovery and potential impeachment on cross-examination. That policy is a far more reasonable (and constitutional) way of handling the risk of potentially unfavorable written opinions from experts than simply avoiding experts altogether.
In short, the decision not to call a psychiatrist was not the result of reasoned strategic decisions, no matter how many illogical post-hoc rationalizations the State and the majority may advance. See Wiggins, 539 U.S. at 526-27, 123 S.Ct. 2527 (“The ‘strategic decision’ the state courts and respondents all invoke to justify counsel’s limited pursuit of mitigating evidence resembles more a post-hoc rationalization of counsel’s conduct than an accurate description of their deliberations prior to sentencing.”) Defense counsel’s failure to call an expert in psychiatry was, as the Supreme Court once put it, “the result of in attention, not reasoned strategic judgment.” Id. at 534, 123 S.Ct. 2527. Consequently, I would hold that Brown’s defense attorneys’ performance was deficient in this respect.
I would also hold that Brown was prejudiced by his counsel’s performance. The lack of expert testimony from a psychiatrist had a devastating impact on Brown’s mitigation efforts at the penalty phase. In the defense’s opening statement, Cleven told the jurors that Dr. Maiuro would explain to them how Brown’s disorder was “not properly diagnosed or properly treated.” However, after the State’s highly damaging cross-examination of Dr. Maiuro, and the testimony of Dr. Brinkley, in which he testified that Brown did not have a mental disorder that required lithium treatment, and without an expert psychiatrist of his own, Mulligan had no choice but to concede in his closing argument that the defense had not shown that Brown should have been treated with lithium, the only treatment it had suggested.
Moreover, during his closing argument, the prosecutor was able to undermine Dr. Maiuro’s testimony by arguing:
Dr. Maiuro really didn’t have enough information to be able to talk about whether the defendant should or should not have been on lithium ... Dr. Maiuro is not a medical doctor, so he does not prescribe lithium to people. He does not have to make the medical analysis of whether lithium is appropriate. That’s not part of his role as a psychologist. So he was kind of out on a limb in that part of his testimony. And I think you recognized that once you heard from Dr. Brinkley.
Dr. Scher’s testimony at the evidentiary hearing in the habeas proceeding demonstrates how helpful a properly qualified psychiatrist would have been for the defense: such a witness would have corroborated Dr. Maiuro’s testimony that Brown suffered from a serious mood disorder and rebutted Dr. Brinkley’s conclusions regarding lithium. As to the first, Dr. Scher testified that based on her personal inter*1044view of Brown and review of his records, Brown suffered from bipolar disorder. She also testified that Schick and Dr. Maiuro’s diagnoses of Brown as manic were consistent with her diagnosis. As to the second, Dr. Scher testified that if Brown had been taking lithium prior to the murder, it “would have controlled the degree of impulsivity and the amount of stimulation he had.” Dr. Scher testified that it was her opinion that if Brown had been on lithium and properly treated, he would have been “less likely” to have committed the murder.
Dr. Scher’s testimony also could have been used to impeach Dr. Brinkley because she testified that psychiatrists are supposed to do a personal interview before making a diagnosis and noted that Dr. Brinkley had failed to do so. Dr. Scher also testified that Dr. Brinkley unreasonably dismissed the fact that prior treatment providers had diagnosed Brown as manic depressive.3
The majority argues that defense testimony by an expert psychiatrist would not have significantly changed the mitigation case because Dr. Brinkley’s testimony did not undermine the defense’s central theory and did not directly contradict Dr. Mauiro’s diagnosis. Maj. Op. at 1034. As explained above, however, this is simply not correct. Dr. Brinkley’s unchallenged testimony cast doubt over the entire mental illness argument by undermining Dr. Mauiro’s diagnosis and eliminating from consideration the lithium issue, which was the most tangible way for the jury to understand the magnitude of Brown’s illness. Dr. Brinkley did not have to contradict Dr. Mauiro’s diagnosis directly or challenge any of the testimony about Brown’s family background: the doubts he raised about Brown’s mental disorder and his undisputed testimony about lithium were enough to devastate the mitigation theory.
In this case, defense counsel’s failure to call a psychiatrist left the jury with no opportunity to consider the role that medical treatment with lithium could have played in controlling Brown’s behavior, nor did the jurors hear any contradiction of Dr. Brinkley’s devastating testimony that Brown did not have any disorder that could be treated with lithium. There is a reasonable probability that, had the jury heard the testimony of a psychiatrist, like Dr. Scher, who would have testified that Brown did have a mental disorder and that lithium treatment would have made him less likely to commit the crime with which he was charged, at least one juror would have concluded that Brown’s treatable disorder was a mitigating circumstance that warranted a life sentence rather than death under Washington law.
II. Failure to Call Schick
Schick is an experienced professional counselor who is licensed to make psychological diagnoses and who observed and treated Brown at the Oregon state prison for over two years. Of all the potential witnesses who were available to the defense at the time of trial, she was the only one who had witnessed the impact of lithium treatment on Brown. Nevertheless, none of Brown’s attorneys — only an investigator — interviewed Schick and Schick was never asked to testify at the penalty phase.
*1045At the evidentiary hearing, Schick testified that she had initially diagnosed Brown with bipolar disorder, but upon further observation revised her diagnosis to unipolar disorder because Brown did not appear to experience the severe depressive episodes that are characteristic of bipolar disorder. She requested a lithium trial to stabilize Brown’s mood disorder and noticed an improvement in the month that she observed him on the drug. According to Schick, Brown “was able to focus longer and less likely to preoccupy himself with some of those grandiose thoughts.” In her declaration, Schick stated that she “believe[s] that this tragedy might have been averted if Brown’s mental health treatment had been started sooner and his medication monitored more carefully.”
The majority claims that counsel made a reasoned decision not to call Schick because her qualifications were not as impressive as those of Dr. Maiuro, and because she, like Dr. Maiuro, could not prescribe lithium. Maj. Op. at 1035-36. Again, the majority’s rationalizations for defense counsel’s failings are not supported by the record. In fact, the failure to call Schick was a result of inattention, not strategy.
Lauren Sonik, a defense investigator, spoke with Schick a few months before trial began. None of Brown’s attorneys, however, ever spoke with her. Mulligan did not remember making a “specific tactical decision not to call Ms. Schick as a witness at the penalty phase.” Cleven testified that he and his colleagues never “interviewed her closely as a potential witness” and could not “recall what considerations there were in deciding not to call her as a witness.” Hupp remembers “positive things” about Schick and remembers that she was exactly the sort of witness with the kind of information the defense was looking for. Hupp, like the others, did not remember why they did not call her. Thus, the majority’s speculative explanation for defense counsel’s failure to call Schick to testify has no basis in fact and is not supported by the record or the recollections of the attorneys.
Moreover, the majority’s post-hoc rationalizations are unreasonable. First, the argument that Dr. Maiuro was better suited to testify because he has more degrees and professional accomplishments misses the point. Schick should have been called also, not because of her long curriculum vitae, but rather because of her unique experience with Brown. She was the only counselor who frequently observed and treated Brown over a period of time, observed his progress on lithium, and the only witness who could have projected the effect of further treatment. Thus, she was the best person to testify about Brown’s responsiveness to medical treatment. Despite the fact that Schick’s notes were read to the jury by Dr. Maiuro, the scant observations contained in them were no substitute for the detailed testimony that Schick could have given about her personal experience with Brown.
The majority’s second explanation — that Schick was not a medical doctor and thus not qualified to prescribe lithium — is also contrary to reason and logic. First, that was clearly not counsel’s actual reason for not calling Schick, given that it did not deter them from calling Dr. Maiuro. More important, however, Schick would not have been called to testify about whether Brown should have been prescribed lithium. As explained above, she would have been asked about the change in Brown’s behavior once he began taking the drug. Her ability to prescribe medication has nothing to do with her ability to observe and evaluate a patient who is on medication.
Brown’s counsel’s failure to call Schick was not the result of a reasoned tactical decision; it was, rather, another failure properly to prepare the case for trial. Al*1046though Schick could have been a critical witness, no lawyer followed up on the investigation and interviewed her to determine how valuable her testimony might be. Thus, it would have been impossible for counsel to make an informed strategic decision not to call her. Counsel’s representation fell below professional standards when they decided not to call Schick to testify at the penalty phase without interviewing her and without sufficient information to make an informed decision. Accordingly, I would hold that this decision also constituted deficient performance.
Schick’s testimony would have corroborated Dr. Maiuro’s testimony and directly rebutted Dr. Brinkley’s claim that there was nothing in Brown’s record that suggested he suffered from a mental disorder that could be treated with lithium. Unlike Dr. Maiuro or Dr. Brinkley, Schick could have provided first-hand testimony about the effect of lithium on Brown.
Because she did not testify, the jury heard about Schick only through the testimony of Dr. Maiuro and Dr. Brinkley.
Dr. Maiuro testified that, according to the prison records, Schick concluded that the treatment program resulted in “no change” in Brown. Dr. Brinkley also testified that Schick “said in her summary statement of [Brown’s] course of treatment with her that she saw no change.”
The record reveals, however, that had the defense called her to testify, Schick would have told the jury that Brown improved on lithium. In her declaration, filed in the habeas proceeding, Schick stated that she had “noticed a marked improvement in Brown’s symptoms” once his lithium treatment began in Oregon state prison. She noted that Brown’s “internal energy was immediately decreased” and that he “was able to think and plan better.” Schick also declared that, had she been called to testify, she would have stated that it was her “firm professional belief that Brown had a major psychiatric disorder that was never effectively treated” and that “this tragedy might have been averted” if Brown’s medication had been properly administered and monitored.
At the evidentiary hearing, Schick testified that during the month Brown was on lithium she observed “a gradual change” as Brown was “less likely to preoccupy himself with some of those grandiose thoughts.” She also testified that she saw some indication of the remission of Brown’s symptoms after he was placed on the lithium trial. Thus, her testimony could have impeached Dr. Brinkley’s conclusion that Schick had observed no change in Brown as a result of his lithium treatment.4 I would hold that, standing alone, counsel’s failure to call Schick prejudiced Brown, or, at the very least, that he was prejudiced when that failure is considered in conjunction with counsel’s failure to call a psychiatrist as an expert witness.
III. Failure to Cross-Examine Dr. Brinkley
Brown’s defense lawyers elected not to cross-examine Dr. Brinkley after he testified that Brown did not have a mental *1047disorder for which lithium could be prescribed. As a result, the jury never heard that Dr. Brinkley made his evaluation and rendered his opinion without ever interviewing Brown, Brown’s family, or any of the doctors and therapists who had treated him in the past. Moreover, the jury never heard that Dr. Brinkley’s review of Brown’s records was incomplete and that he had no knowledge of many aspects of Brown’s background that are significant to diagnosing mood disorders.
The majority claims that Brown’s attorneys made a tactical decision not to cross-examine Dr. Brinkley. Maj. Op. at 1036. That a decision can be labeled “tactical,” however, does not end the Strickland inquiry. Rather, “a reviewing court must consider the reasonableness of the investigation said to support that strategy.” Wiggins, 539 U.S. at 527, 123 S.Ct. 2527 (citing Strickland, 466 U.S. at 691, 104 S.Ct. 2052). Counsel performs inadequately at the penalty phase when he makes a tactical decision based upon inadequate investigation. See Wiggins, 539 U.S. at 536, 123 S.Ct. 2527. Here, there is no evidence that the decision not to cross-examine Dr. Brinkley was based on a reasonable investigation.
Cleven testified that he was prepared for the cross-examination and thought that it was important for the jury to hear concessions from Dr. Brinkley about the limited information upon which he based his testimony. During a break in the testimony, however, Mulligan, the lead attorney, ordered him not to conduct cross-examination. Cleven reported that Mulligan told him that “he had seen cross-examinations backfire” and that the points that Cleven wanted to get out in cross-examination could be made in the closing argument.
Mulligan admitted that he played no role in preparing to cross-examine Dr. Brinkley because the doctor was eleven’s witness, nor did he attend the short interview that the other two defense attorneys conducted with Dr. Brinkley on the morning of his testimony. Thus, Mulligan did not know in advance that Dr. Brinkley would testify that Brown did not have a disorder for which lithium should be prescribed and that Brown did not benefit from his lithium treatment at the Oregon state prison. If, like Cleven, Mulligan had played a role in preparing for cross-examination and if he, like Cleven, had been present at the interview with Dr. Brinkley, he likely would have come to the same conclusion as Cleven — that cross-examining Dr. Brinkley “would not pose any risk to [the case] at all” and indeed would likely result in answers that would “seriously call his testimony into question.” Additionally, Mulligan would have known that the points that Cleven wanted to make on cross-examination, namely that Dr. Brinkley made his assessment of Brown without interviewing him personally and on the basis of an incomplete review of Brown’s record, could not be made during the closing argument without first establishing their factual basis by questioning Dr. Brinkley. Instead, Mulligan made an uninformed, on-the-spot decision and forbade Cleven from conducting the cross-examination that might have helped save Brown’s life. As Cleven testified at the evidentiary hearing, the decision not to cross-examine Dr. Brinkley was not an informed one and it was not based on reasonable consideration of its merits and possible drawbacks. We need not defer to Mulligan’s hasty decision because it was made without the benefit of adequate investigation and thus fell below prevailing professional standards. I would hold that the decision not to cross-examine Dr. Brinkley constituted deficient performance.
Dr. Brinkley’s testimony was quite prejudicial to Brown’s mitigation case. At trial, Dr. Brinkley testified that he did not see anything in the Oregon state prison records that explained why lithium was *1048prescribed for Brown and that he “had no clear indication from the records [he] reviewed that [Brown] had a disorder for which lithium was appropriate.”
As a result of defense counsel’s failure to cross-examine Dr. Brinkley, the prosecutor was able to argue in his closing argument that:
Apparently, the defense didn’t dispute any of the findings of Dr. Brinkley. They never asked him questions. They apparently were satisfied with his expertise and his findings.
The habeas evidentiary hearing confirms that if Cleven had cross-examined Dr. Brinkley, the jury would have heard that Dr. Brinkley never conducted an interview with Brown and that he did not review all of the records about Brown’s life history that Dr. Maiuro reviewed. In Dr. Brinkley’s deposition, taken prior to the evidentiary hearing, Brown’s counsel asked him about several pieces of information from Brown’s family history that Dr. Brinkley did not review prior to testifying at trial. Dr. Brinkley acknowledged that some of the information from Brown’s childhood that he did not review was “potentially relevant” to the diagnostic question of whether he suffered from a mood disorder. He stated that had he been aware that Brown’s mother had a mood disorder, it would have been relevant to his diagnosis. He admitted that he did not know that Brown started seeing a psychiatrist at age seven, and that his evaluation would have been “more complete” if he had seen the records of the psychiatrist who treated Brown as a child. He also said he did not know that Brown had been diagnosed with cyclothymic disorder in 1982 and that it “would have been of interest” to him since cyclothymic disorder may precede bipolar disorder. Moreover, he testified that the behavioral patterns Schick observed in Brown at the state prison are “consistent” with a mood disorder. This testimony all demonstrates that cross-examination would have revealed to the jury that Dr. Brinkley’s assessment was based upon an incomplete review of Brown’s relevant mental health history and family background.
The majority claims that cross-examination of Dr. Brinkley would have backfired because he testified at the evidentiary hearing that Brown did not have manic syndrome, but instead was a sociopath. However, there is no suggestion that defense counsel intended to ask Dr. Brinkley for his diagnosis of Brown on cross-examination. Cleven planned to ask Dr. Brinkley whether he interviewed Brown and whether he reviewed all of the relevant records. Moreover, Dr. Brinkley’s testimony on cross-examination would not have damaged Brown any more than his testimony on direct. On direct, he explained at length that lithium was used to treat mood disorders including bipolar disorder and manic symptoms. He then testified that Brown did not suffer from a disorder that could be treated with lithium. The conclusion that the jury naturally reached using elementary logic is that Dr. Brinkley believed that Brown did not have manic syndrome or any other mood disorder. Dr. Brinkley was an experienced witness. Had he wanted to say explicitly that Brown did not have any mood disorder or had he been willing to offer a different diagnosis, he surely would have done so on direct. There was no need to wait for cross or redirect.5
*1049The majority also argues that had he been cross-examined, Dr. Brinkley would have revealed that defense counsel were challenging his authority to interview Brown. Maj. Op. 1036-37. First, this contradicts the majority’s other claim that Dr. Brinkley would have said that he had no interest in interviewing Brown because Brown was a liar. Maj. Op. at 1036-37. The majority cannot have it both ways. Second, the record does not indicate that defense counsel refused to permit an interview with Brown. The defense objected to the State’s attempt to have Dr. Brinkley observe Brown’s demeanor in the courtroom, but it invited the State to make a motion allowing Dr. Brinkley to interview him. The State failed to do so.
It is more likely that the majority’s other explanation for Dr. Brinkley’s failure to interview Brown is accurate — that Dr. Brinkley did not think it was necessary because he thought that Brown was a liar. Maj. Op. 1036-37. This is a puzzling excuse for Dr. Brinkley to affirm given that he relied solely on the notes of other professionals who interviewed Brown. Perhaps he believed them to be better at detecting lies than he. Regardless, even if Dr. Brinkley denied the value of a personal interview, it does not mean that the jury’s assessment of his testimony would not have been affected by learning that he did not interview Brown. Moreover, the evidentiary hearing revealed that cross-examination would have enabled the defense to elicit from Dr. Brinkley the concession that he thinks clinical psychologists are “competent to make psychiatric diagnoses” and that he has worked with Dr. Maiuro in the past and considers him to be “a competent practitioner.” This concession would have mitigated the damage done by the State’s cross-examination of Dr. Maiuro in which the psychologist was forced to admit that he was not qualified to prescribe lithium and by the State’s closing argument in which the prosecutor commented that Dr. Maiuro was “out on a limb.”
There is a reasonable probability that, had the jury heard that Dr. Brinkley never interviewed Brown, did not review the same records as did Dr. Maiuro, and believed that Dr. Maiuro was competent to render an opinion, at least one juror would have given far greater weight to the opinions of Dr. Maiuro with respect to the question whether Brown’s mental disorder and the lack of lithium should be considered to constitute a mitigating factor.
IV. Cumulative Error
The three aspects of counsel’s deficient performance, viewed together, should undermine one’s confidence in the outcome of the penalty phase. Multiple errors, even if harmless individually, may entitle a petitioner to habeas relief if their cumulative effect prejudiced the defendant. Mak v. Blodgett, 970 F.2d 614, 622 (9th Cir.1992), cert. denied, 507 U.S. 951, 113 S.Ct. 1363, 122 L.Ed.2d 742 (1993); see also United States v. Tucker, 716 F.2d 576, 595 (9th Cir.1983); Cooper v. Fitzharris, 586 F.2d 1325, 1333 (9th Cir.1978) (en banc) (holding that “prejudice may result from the cumulative impact of multiple deficiencies”).
Defense counsel’s errors resulted in the failure adequately to present Brown’s strongest mitigation argument. The failure to call a psychiatrist to testify about Brown’s mental disorder and his need for lithium, the failure to call Schick to testify about her observations while treating Brown over a period of time, including her observations about Brown’s responsiveness to lithium, and the failure to cross-examine Dr. Brinkley after his devastating testimo*1050ny that Brown did not have a disorder for which lithium treatment was appropriate, likely led the jury to believe that Brown’s mental disorder was not serious and, consequently, did not constitute a mitigating factor that warranted a life sentence. Had counsel for the defense performed competently, they could have presented a far stronger affirmative case for the existence of Brown’s mental disorder and the effectiveness of lithium treatment, as well as undermined the State’s expert witness’s contrary opinions. Properly presented, evidence that Brown suffered from a mental disorder that contributed to the murder but that could have been treated with lithium, could have tended to establish a powerful mitigating factor. Instead, defense counsel failed to present an adequate case and counsel left themselves vulnerable to the damaging cross-examination of Dr. Maiuro, testimony of Dr. Brinkley, and closing argument by the State.
These are not errors that “had an isolated, trivial effect.” Strickland, 466 U.S. at 695-96, 104 S.Ct. 2052. Had the jury been able to place Brown’s mental illness on the mitigating side of the scale, there is a “reasonable probability” that, at least one juror would have concluded that “the balance of aggravating and mitigating circumstances did not warrant death.” Id. at 695, 104 S.Ct. 2052. As the Supreme Court has held:
[Although we suppose it is possible that a jury could have heard [ ] all [the mitigation evidence] and still have decided on the death penalty, that is not the test. It goes without saying that the undiscovered mitigating evidence, taken as a whole, might well have influenced the jury’s appraisal of [the defendant’s] culpability.
Rompilla v. Beard, 545 U.S. 374, 393, 125 S.Ct. 2456, 162 L.Ed.2d 360 (2005) (quoting Wiggins, 539 U.S. at 538, 123 S.Ct. 2527) (internal quotation marks omitted).
I would hold that defense counsel’s three serious errors — failure to hire a defense expert psychiatrist, failure to call Schick to testify, and failure to cross-examine Dr. Brinkley — prejudiced Brown. He received ineffective assistance of counsel in violation of the Sixth Amendment and, as a result, the writ of habeas corpus should issue.
V. Conclusion
I respectfully dissent. I would grant the writ of habeas corpus and vacate Brown’s death sentence.
. American Bar Association Guidelines for the Appointment and Performance of Counsel in Death Penalty Cases, 11.41(C) (1989), http:// www.abaneVorg/death penalty/resources/docs/1989Guidelines.pdf.
. Defense counsel considered seeking a continuance for unrelated reasons at one point during the trial, but the decision to obtain Dr. Brinkley’s services (or those of some other psychiatrist) did not prompt them to consider one.
. We need not belabor the point that the majority raises of whether Dr. Brinldey was more qualified than Dr. Scher because we have no way of knowing which psychiatrist defense counsel would have found. Maj. Op. at 1035. Even if Dr. Brinkley had a longer curriculum vitae than the defense's psychiatrist, the fact that the defense expert would have interviewed Brown would have more than made up for whatever credential deficit there may have been.
. It appears that the basis for Dr. Maiuro and Dr. Brinkley's testimony that Schick did not notice any change in Brown on lithium was a single checked box on a case summary report dated June, 1993. This was three years after Schick stopped seeing Brown. According to the defense investigator's notes, Schick was leaving the prison at that time but before she left she had to go through all of her old files and close all of her cases. This is probably why she completed the report about Brown at such a late date. Although she did check the box that read "no change,” her notes in the report indicate that the lithium was working, which is consistent with Schick's declaration and testimony at the evidentiary hearing. Schick’s testimony would have dispelled any doubts that might have been raised by the hastily checked "no change” box in the report.
. The majority also notes that at oral argument, counsel conceded that had Dr. Brinkley testified as he did in his habeas deposition, it would have been unhelpful to Brown. Maj. Op. at 1036. However, there is no indication that defense counsel intended to ask the questions that would have elicited this unhelpful testimony. Cleven testified that he was prepared to ask Dr. Brinkley whether he interviewed Brown and whether he reviewed all of *1049the relevant background materials before reaching his conclusions. He gave no indication that he would have asked Dr. Brinkley to render a complete diagnosis.