with whom joins, concurring.
ILANA DIAMOND ROVNER, Circuit Judge,I fully agree with the majority’s well-reasoned conclusions that Eddmonds’ conviction and sentence were ultimately not affected by his attorney’s performance. I write separately, however, because I would make the additional finding that defense counsel’s representation during the critical death penalty stage, while not in the end prejudicial, was constitutionally deficient.
As the majority notes, we measure counsel’s conduct during a capital sentencing hearing against the well-established standard that a “significant effort, based on reasonable investigation and logical argument,” must be *1324made to mitigate a client’s punishment. Rabat v. Thieret, 867 F.2d 361, 369 (7th Cir.1989); see also Stewart v. Gramley, 74 F.3d 132, 135 (7th Cir.1996). Although we have recognized that the exact parameters of counsel’s duty to investigate and present mitigating factors may be difficult to ascertain, see Gramley, 74 F.3d at 135, in certain situations “it will be apparent from the evidence concerning the circumstances of the crime, from conversation with the defendant, or from other sources of information not requiring fresh investigation, that the defendant has some mental or other condition” that will warrant further investigation. Id. In such situations, “the failure to investigate will be ineffective assistance.” Id.
In my view, it was apparent from the evidence in the record prior to the sentencing hearing that an investigation by defense counsel to determine if Eddmonds was suffering from an extreme mental and emotional disturbance (“mental disturbance”) at the time of the crime was required. At the time of the penalty phase, the file before defense counsel revealed that Eddmonds had been examined more than ten times to determine his competency to stand trial, a clear indication that his mental stability was seriously questionable. While the evaluations of fitness became increasingly positive over time, the professional opinions rendered closest to the time of the crime (which is the relevant time for a mental disturbance determination) concluded that Eddmonds was mentally unfit to stand trial.1 Additionally, among the numerous psychological and psychiatric reports evaluating Eddmonds’ competency to stand trial, at least four concluded that Eddmonds suffered from some form of schizophrenia, while others reported diagnoses of various borderline and anti-social personality disorders mixed with depressive features.2 None concluded that Eddmonds was without psychological problems. The documents in counsel’s discovery file also revealed that in 1973, while serving his second sentence for rape, Eddmonds was diagnosed with schizophrenia and spent three months in a psychiatric hospital.3 Further, the record contained evidence that during his past encounters with the criminal justice system Eddmonds had attempted suicide several times and engaged in self-mutilation. Thus even a cursory review of Eddmonds’ file would have revealed long-standing, complex, and often severe mental problems. Such a well-documented mental health history would undoubtedly alert the committed attorney to the existence of possible mitigation evidence and clearly invoke the duty of reasonable investigation. We and other circuits have found a duty to inquire further with much less abundant and consistent documentation of mental illness. See, e.g., Brewer v. Aiken, 935 F.2d 850, 857-58 (7th Cir.1991); Antwine v. Delo, 54 F.3d 1357, 1367 (8th Cir.1995); Hill v. Lockhart, 28 F.3d 832, 845 (8th Cir.1994); Stephens v. Kemp, 846 F.2d 642, 652-53 (11th Cir.1988).
The fact that there were two findings of sanity contained in the file before defense counsel does not make his decision to forego further investigation a reasonable one. A finding of sanity is not equivalent to a finding that the defendant was not influenced by a mental disturbance. The Illinois death pen*1325alty statute provides that one mitigating factor to be considered is “whether the murder was committed while defendant was under the influence of extreme mental or emotional disturbance, although not such as to constitute a defense to prosecution.” 720 ILCS 5/9 — 1(e)(2) (emphasis added). A person can therefore be declared sane, yet still be suffering from a mental disturbance. This certainly follows, because if a person were insane, his condition would not simply be a mitigating factor, it would be a shield from a sentence of death, if not a defense to the crime itself. See 720 ILCS 5/6-2 (setting forth insanity defense). For this reason, if a finding of sanity is sufficient to halt further investigation, then there would rarely be a duty to independently investigate possible mental disturbance, since presumably only sane criminals are subject to the death penalty. A determination of sanity therefore cannot be a sufficient reason to forego inquiry into psychological problems for mitigation purposes. See Stephens, 846 F.2d at 653; Loyd v. Whitley, 977 F.2d 149, 156-57 (5th Cir.1992).
Yet Eddmonds’ attorney essentially abdicated his duty to make reasonable inquiry of mitigating circumstances. In fact, he admitted at both the state court post-conviction hearing and the district court hearing that he could not recall taking any investigative steps prior to the death penalty hearing. There is no indication in the record that counsel asked any mental health expert to examine Edd-monds for possible mental disturbance at the time of the murder, nor did he seek to have any of the experts who had recently evaluated Eddmonds for competency and insanity reexamine him in order to render an opinion on mental disturbance. There was apparently no endeavor by counsel to obtain the actual records and reports documenting Edd-monds’ past psychological problems, which were clearly referenced and mentioned in counsel’s discovery file. Counsel also chose not to interview any of Eddmonds’ family members, who may have had (and as it turns out did have) anecdotal information that would have shed light on Eddmonds’ mental condition. It appears evident that defense counsel not only neglected to pursue a meaningful investigation, he attempted, in effect, no investigation.
The district court and the government attempt to cast defense counsel’s “choice” to forego investigation as a strategic maneuver supported by “reasonable professional judgment,” which we generally afford great deference. See Strickland v. Washington, 466 U.S. 668, 689, 690-91, 104 S.Ct. 2052, 2065, 2066, 80 L.Ed.2d 674 (1984) (“[Strategic choices made after less than complete investigation are reasonable to the extent that reasonable professional judgments support the limitations on investigation.”). 466 U.S. at 690-91, 104 S.Ct. at 2066; see also Ant-wine, 54 F.3d at 1367. It is contended that counsel did not seek an evaluation for possible mental disturbance because he was concerned that investigation might result in conclusive proof that Eddmonds was not suffering from a mental disturbance, and then counsel would have been unable to “suggest” that Eddmonds had acted under a mental disturbance through Dr. Reifman’s testimony. I am unable to accept this “strategy” as supporting a decision to forsake investigation in this case.
First, counsel never claimed that this creative reason was his rationale for not pursuing mitigating evidence, nor did he follow this strategy, i.e., he never actually “suggested” mental disturbance to the sentencing judge. And we are constrained from “construct[ing] strategic defenses which counsel does not offer.” Harris v. Reed, 894 F.2d 871, 874 (7th Cir.1990); cf. Strickland, 466 U.S. at 673, 699-700, 104 S.Ct. at 2057, 2070-71 (accepting strategy that counsel stated, explained, and clearly followed); Burger v. Kemp, 483 U.S. 776, 790-95, 107 S.Ct. 3114, 3123-26, 97 L.Ed.2d 638 (1987) (accepting reasoned explanations offered by counsel for each decision).
Second, and more importantly, this is not a case where counsel’s conversations with the defendant or a preliminary investigation had “given counsel reason to believe that pursuing [or continuing] certain investigations would be fruitless or even harmful.” Strickland, 466 U.S. at 691, 104 S.Ct. at 2066. In such a case, it may be a reasonable professional judgment to limit investigation and *1326rely only on suggestions or inferences from evidence already in hand. See, e.g., Strickland, 466 U.S. at 691, 699, 104 S.Ct. at 2066, 2070; Burger, 483 U.S. at 796, 107 S.Ct. at 3126. As just observed, however, in this case a cursory discussion with Eddmonds or his family regarding his mental history, or any review of his file, would by no means have led to the judgment that investigation would be fruitless; but rather should have spurred inquiry. In other words, on the record before him, defense counsel had no solid or reasonable basis for harboring a fear that new reports on mental disturbance would be conclusively negative.4 Further, even if the evaluations had conclusively reported that Eddmonds was not influenced by a mental disturbance at the time of the crime, then at that point counsel could have made an informed decision not to present the testimony, and the negative reports could have remained confidential. See People v. Knuckles, 165 Ill.2d 125, 140, 209 Ill.Dec. 1, 650 N.E.2d 974 (1995) (defense psychiatric expert’s conclusions privileged unless called to testify). Thus, the possibility of uncovering harmful evidence did not justify totally foregoing investigation in this case. Full and complete investigation is absolutely crucial in a death penalty situation, and on a record such as this, counsel must be encouraged to fulfill this duty, rather than allowed to circumvent the obligation based on a “fear of what I may find” theory. In my opinion, counsel’s decision not to initiate an investigation that was clearly called for by the evidence was not supported by reasonable professional judgment. Therefore, his actions should not be shielded from a label of deficiency under the rubric of strategy.5
Moreover, defense counsel did not even . effectively “suggest” the possibility of mental disturbance, as the government and the district court claimed it was his “strategy” to do. He did not present the numerous reports demonstrating Eddmonds’ history of mental problems (which were readily available at the time of sentencing), nor did he submit the reports into evidence. Counsel did put Dr. Reifman on the stand and elicit testimony from him based on his 1977 report that at one point Eddmonds had been diagnosed as schizophrenic and that his condition had caused him to be out of touch with reality and incompetent to stand trial. Counsel did nothing, however, to prepare Dr. Reifman for his testimony. He did not ask Dr. Reifman to review his own reports on Eddmonds, let alone those of others concerning Eddmonds’ mental problems, nor did he ask Dr. Reifman to reevaluate Eddmonds. Counsel simply asked the doctor to bring his file to court. Dr. Reifman, therefore, had no knowledge of the four other schizophrenia diagnoses or of Eddmonds’ well-documented, long-standing history of mental illness. Additionally, because counsel had proceeded no further in this regard, he was unable to offer any new reports or testimony supporting Dr. Reifman’s prior schizophrenia diagnosis.
In this situation it is not surprising that Dr. Reifman’s testimony was severely impeached and undermined upon cross examination. The government forced Dr. Reifman to admit that his conclusions were not verified by any other source, that the stress of incarceration for murder may have induced Eddmonds’ mental condition at the time of Dr. Reifman’s interview, that other evaluators and he himself at a later date had concluded that Eddmonds was fit for trial, and finally that he was unable to render an opinion on the critical question of whether Eddmonds was influenced by a mental dis*1327turbance at the time of the offense (since he had not been asked to evaluate Eddmonds for such a purpose). Defense counsel did not seek to rehabilitate Dr. Reifman with the reports that were readily accessible. In fact, he performed no redirect and rested his case following this damaging cross examination, thereby abandoning the witness and the modest attempt at mitigation. The ultimate import of Dr. Reifman’s testimony was that Eddmonds’ psychotic condition at the time of Dr. Reifman’s 1977 interview was a one-time aberration, and thus the mitigating value of the testimony was de minimus, if non-existent. Counsel did not even mention Dr. Reif-man’s conclusions in his short closing argument. Had Dr. Reifman been adequately prepared, supported, and rehabilitated, the effort on counsel’s part to fulfill his mitigation duties could be viewed more approvingly. The manner in which the witness was managed, however, rendered Dr. Reifman’s testimony possibly more damaging than helpful and only supports a finding of deficient performance.
Given counsel’s ill-considered decision to forsake investigation, his failure to present the evidence he already possessed, and his questionable handling of Dr. Reifman’s testimony, I can only conclude that defense counsel’s representation at sentencing “fell below an objective standard of reasonableness,” Brewer, 935 F.2d at 858, and therefore was “outside the wide range of professionally competent assistance.” Strickland, 466 U.S. at 690, 104 S.Ct. at 2066. As we have previously stated:
[DJefense counsel must make a “significant effort,” based on reasonable investigation and logical argument to ably present the defendant’s fate to the [judge] and to focus the attention of the [judge] on any mitigating factors.... [C]ounsel may not treat the sentencing phase as nothing more than a mere post-script to the trial. While the Strickland threshold of professional competence is admittedly low, the defendant’s life hangs in the balance at a capital sentencing hearing. Indeed, in some cases, this may be the stage of the proceedings where counsel can do his or her client the most good.
Kubat, 867 F.2d at 369. Counsel’s performance in this case did not constitute the “significant effort” at mitigation this court prescribed in Kubat.
In the end, however, I must agree with the majority that defense counsel’s deficient performance did not ultimately prejudice Edd-monds. We now know that the results of the investigation that should have been conducted would have rendered conflicting results, and I cannot say that there is a reasonable probability that such inconclusive results would have been sufficient to overcome Edd-monds’ serious criminal history and the horrifying nature of Richard Miller’s death. This concurrence is proffered mainly to express deep concern over the absence of a meaningful investigation in preparation for the mitigation phase of this matter and in order to underscore that such a subminimum performance should not be viewed as an acceptable benchmark of representation in a capital case.
. The first doctor to evaluate Eddmonds for competency, Dr. Reifman, found Eddmonds unfit to stand trial. On the same day Dr. Rabin found his competence questionable. Several weeks later, upon initial examination, Dr. Goldsmith of the Illinois State Psychiatric Institute found Edd-monds unfit to stand trial. Dr. Goldsmith eventually concluded, however, that Eddmonds was fit for trial. Near the same time Dr. Stipes concluded Eddmonds was unfit.
. In December of 1977, Dr. Reifman diagnosed Eddmonds as suffering from undifferentiated schizophrenia. On the same day, Dr. Rabin concluded that Eddmonds suffered from paranoid schizophrenia. Dr. Stipes reached the same diagnosis of paranoid schizophrenia in January of 1978. In July of 1978, Dr. Kaplan concluded that Eddmonds suffered from undifferentiated schizophrenia, although he later changed his diagnosis to antisocial personality with depressive features, as did Dr. Rabin. This is not an exhaustive list of the reports documenting Edd-monds' mental problems.
.Dr. C.T. CiaTteo opined that Eddmonds was in an early schizophrenic process in February of 1973. In March of the same year Eddmonds was diagnosed by several psychologists with borderline personality disorder. These diagnoses were made during Eddmonds’ three month stay at Menard Psychiatric Hospital.
. And, as we now know, Drs. Conroe and Stipes would have testified that Eddmonds was indeed influenced by a mental disturbance at the time of the offense.
. The district court also seemed to suggest that counsel’s decision was strategically reasonable because the record now reflects that any investigation would have yielded only conflicting reports. The government made a similar argument in defense of counsel's "choice," contending that ”[c]ounsel cannot be faulted for failing to present evidence that would not have been helpful to petitioner's case.” However, both of these arguments are inconsistent with the principle, recognized by both the district court and the government, that counsel's actions must be evaluated in light of the evidence available at the time of the contested actions. When defense counsel decided to forego further inquiry, he did not know that conflicting reports would be the result, and therefore this cannot justify his actions. Hindsight is relevant only to the question of prejudice, not to that of deficient performance.