RHESA HAWKINS BARKSDALE, Circuit Judge:
Louisiana challenges the habeas relief granted Leo Wilson on his state conviction for armed robbery, the issue being whether the prosecution violated Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963), by not disclosing a police report that could have been used to impeach the credibility of one of the two victims ■ (Leonard Pierce), the sole witness to definitely identify Wilson. The determinative question is whether the report was “material”; this is, whether “there is a reasonable probability that, had the [report] been disclosed to [Wilson], the result of the [jury trial] would have been different.” United States v. Bagley, 473 U.S. 667, 682, 105 S.Ct. 3375, 3383, 87 L.Ed.2d 481 (1985). This case presents a close call; but, because we conclude that there is not a reasonable probability that the result of the trial would have changed, we REVERSE and REMAND.
I.
In March 1983, a jury convicted Wilson for the armed robbery in 1982 of Pierce and Charles Bowie. He was sentenced to two concurrent 50-year terms of imprisonment at hard labor, without the benefit of probation, parole, or suspension of sentence. The conviction was affirmed on direct appeal. State v. Wilson, 463 So.2d 655, 656 (La.Ct.App. 4th Cir.1985), writ denied, 466 So.2d 466 (La.1985).
After exhausting state remedies, State v. Wilson, 587 So.2d 691 (La.1991), Wilson sought federal habeas relief, claiming, inter alia, that his conviction was obtained in violation of due process because the prosecution suppressed material evidence (police report) favorable to his defense. After an evidentia-ry hearing, the magistrate judge found that the prosecution had not disclosed the report, which included Pierce’s description of the robbery to the investigating officers. And, after comparing the versions of the robbery presented in the report and in Pierce’s trial testimony, the magistrate judge found that they differed in material respects, and recommended that relief be granted pursuant to the due process claim.2 In a most thorough opinion, the district court adopted the recommendation and granted habeas relief.
[435]*435II.
In Brady v. Maryland, the Supreme Court held that “the suppression by the prosecution of evidence favorable to an accused upon request violates due process where the evidence is material either to guilt or punishment, irrespective of the good faith or bad faith of the prosecution”. 373 U.S. at 87, 83 S.Ct. at 1196-97. “The guiding principle of Brady is that a jury should be permitted to hear and evaluate all relevant evidence going to a defendant’s guilt or punishment”. Fulford v. Maggio, 692 F.2d 354, 357 (5th Cir.1982), rev’d in part on other grounds, 462 U.S. 111, 103 S.Ct. 2261, 76 L.Ed.2d 794 (1983). As stated in United States v. Bagley:
The Brady rule is based on the requirement of due process. Its purpose is not to displace the adversary system as the primary means by which truth is uncovered, but to ensure that a miscarriage of justice does not occur. Thus, the prosecutor is not required to deliver his entire file to defense counsel, but only to disclose evidence favorable to the accused that, if suppressed, would deprive the defendant of a fair trial.
473 U.S. at 675, 105 S.Ct. at 3379-80 (footnotes omitted).
For obvious due process (fair trial) reasons, impeachment evidence, as in issue here, is covered by Brady. United States v. Bagley, 473 U.S. at 676, 105 S.Ct. at 3380. The Court had earlier held in Giglio v. United States, 405 U.S. 150, 92 S.Ct. 763, 31 L.Ed.2d 104 (1972), that, “[wjhen the ‘reliability of a given witness may well be determinative of guilt or innocence,’ nondisclosure of evidence affecting credibility falls within [Brady’s] general rule”. Id. at 154, 92 S.Ct. at 766 (quoting Napue v. Illinois, 360 U.S. 264, 269, 79 S.Ct. 1173, 1177, 3 L.Ed.2d 1217 (1959)).
To prevail under Brady, Wilson must show that (1) the report was not disclosed,3 (2) it contained evidence favorable to his defense, and (3) that evidence was material. United States v. Sink, 586 F.2d 1041, 1051 (5th Cir.1978), cert. denied, 443 U.S. 912, 99 S.Ct. 3102, 61 L.Ed.2d 876 (1979). The first factor is not in issue.4
A.
In order to determine whether the report contained evidence favorable to Wilson, it is necessary to contrast, in detail, the report and Pierce’s trial testimony. The report (narrative section) describes the robbery as follows:
Leonard Pierce stated [to] reporting officers M. Stewart and R. Monteverde that on 9-10-82/4:30 PM he and Charles Bowie [436]*436were walking lake bound on A.P. Tureaud toward Broad St. Upon reaching N. Broad St. and A.P. Tureaud he noticed 2 unk blk male behind him. L. Pierce didn’t pay any attention to them and continued home down Florida Ave. Upon reaching Florida Ave. and A.P. Tureaud L. Pierce saw a third unk NM come from behind the pumping station toward he and C. Bowie. L. Pierce also states that one first two unk NM came from behind and demand money from he and C. Bowie with 4" folding buck knife. They refused. Then the first unk NM, with knife struck L. Pierce in face and struggle [ejnsued. L. Pierce was then forced on ground by NM # 1 who jabbed L. Pierce in lower back and buttock telling him to stay [on the] ground and to give up the money. Wanted Subject # 1 then went into L. Pierce’s back pocket removing his wallet and taking 1 $5.00 bill from his wallet.
Leonard Pierce also stated that w[h]il[e] this was going on wanted subject # 2 took $6.00 from Charles Bowie and the third wanted subject stood watch. All three wanted subjects heard a car coming and fled down Florida to Broad St. then unknown. Mr. Pierce and Bowie then tried to follow them but no avail. Mr. L. Pierce then continued home where he phoned the police. He also noticed that his pants were cut and it was then he noticed a small incision made by wanted subject # 1 in his lower left buttock....
Charles Bowie' was contacted by phone by Officer M. Stewart and confirmed the above statement by L. Pierce. Mr. C. Bowie was not injured during the armed robbery.
As discussed infra, of critical importance is the fact that Pierce did not sign the report, and there is no evidence that he otherwise adopted the narrative as his own statement. (Pierce did not testify at the federal eviden-tiary hearing.)
At trial, Pierce gave the following account of the robbery:
I was going down A.P. Touro. At the end of A.P. Touro and Broad, I normally go behind the pumping station across the railroad tracks.... As we were walking around the pumping station, ... a field was on my left and the pumping station was on my right. A guy cut across the field. I saw another guy come from the pumping station and another one came from behind me with a knife. The guy with the knife put the knife to my back and told me to give him my money and I refused and the guy that cut across the field came up to me and he tried to go into my pockets and I also stopped that.... I had one of his arms, I broke free and he hit me knocking my glasses off.
Pierce identified Wilson as the person who struck him. He testified that, before his glasses were knocked off, he got a good look at Wilson, and that
[a]fter he knocked my glasses off, I turned to see what had happened, if they had broken or not. When I turned, he [Wilson] grabbed me from behind and wrestled me to the ground.
I tried to get off the ground. That is when the person came up behind me with the knife. He had gone over to my friend and he held him up and he gave him his money and he came back over to me with the knife and the knife was placed in my side. I had to arch my back or I would have been stabbed. They went through my pockets and took the money out and just dropped the wallet. They took the knife and they were pricking me with the knife. He must have hit me one good time and I didn’t even know it, but at that time, that is when I was stabbed. A car came and the two got up and the look-out was already down by the corner and they all ran.
Pierce testified that Wilson was not the one who stabbed him, but instead was the one who was holding him; that Wilson held a knife to his back, however, at some point during the incident; and that he was “absolutely sure” of his identification, and had “no doubt” that Wilson was the person who robbed him, knocked off his glasses, held the knife to his back, and punched him. On cross-examination, Pierce testified that he first saw Wilson when Wilson “cut across the field”. He testified further that the closest [437]*437Wilson got to him was “[f]ace to face”, “less than a foot away from my face”.
To the extent there are discrepancies between the report and Pierce’s testimony, they are favorable to Wilson, because they could have been used to impeach Pierce’s credibility and his identification of Wilson.5 Accordingly, we turn to whether the discrepancies are material.
B.
As noted, the Supreme Cqprt held in United States v. Bagley that, for Brady purposes, it had rejected any distinction between impeachment evidence and other exculpatory evidence. 473 U.S. at 676, 105 S.Ct. at 3380. It reasoned that impeachment evidence is “evidence favorable to an accused”, within the meaning of Brady, “so that, if disclosed and used effectively, it may make the difference between conviction and acquittal”. Id. The Court adopted the following materiality standard for any prosecutorial failure to disclose evidence favorable to the accused:
The evidence is material only if there is a reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different. A ‘reasonable probability’ is a probability sufficient to undermine confidence in the outcome.
Id. at 682, 105 S.Ct. at 3383.
“The question of materiality present in cases in which the accused complains of prosecutorial suppression of material evidence is ... [a] mixed question[ ]' of law and fact calling ultimately for a legal determination”.6 Davis v. Heyd, 479 F.2d 446, 451 (5th Cir.1973); see also Ballinger v. Kerby, 3 F.3d 1371, 1375 (10th Cir.1993)' (“The question of materiality and the possible effect of ... withheld evidence on the verdict[ ] is a mixed question of fact and law”) (internal quotation marks, citation, and brackets omitted); United States v. Rivalta, 925 F.2d 596, 598 (2d [438]*438Cir.), cert. denied, — U.S. —, 112 S.Ct. 215, 116 L.Ed.2d 173 (1991) (same).
1.
“Bagley evidences concern with ‘any adverse [e]ffect that the prosecutor’s failure to respond might have had on the preparation or presentation of the defendant’s case’ ”, Smith v. Black, 904 F.2d at 966 n. 4 (quoting Bagley, 473 U.S. at 683, 105 S.Ct. at 3384 (opinion of Blackmun, J.)). Accordingly, although the Bagley materiality standard applies to a specific request, a general request, or no request at all, “it may be proper to weigh in favor of the accused ‘the more specifically the defense requests certain evidence, thus putting the prosecutor on notice of its value’”. James v. Whitley, 926 F.2d 1433, 1439 (5th Cir.1991) (quoting Bagley, 473 U.S. at 682, 105 S.Ct. at 3383-84 (opinion of Blackmun, J.)).7
Prior to trial, Wilson made only a general request for Brady material (“Motion for Bill of Particulars and Discovery and Inspection”): “Did the State obtain or does the State have any exculpatory evidence or evidence favorable to the defendant and if so, what is the nature and description of such evidence?” The State responded, “None”.
The trial transcript indicates, however, that Wilson’s counsel knew that the report existed. During cross-examination of Officer Bayard, who became involved in the investigation two months after the robbery,8 Wilson’s counsel asked when the report was dated, and who wrote and signed it. Bayard responded that the report was dated September 10, 1982 (the day of the robbery), but could not recall the author’s name. The prosecutor stated that Stewart and Montev-erde were the officers who initially interviewed the victims and prepared the report. At Wilson’s counsel’s request, the trial court ordered the issuance of subpoenas for those officers.
When Wilson’s counsel attempted to cross-examine Bayard about the report’s contents, the court sustained the State’s hearsay objection. In any event, Wilson’s counsel cross-examined both victims about their statements to the police officers on the day of the robbery. During the presentation of defense witnesses, the trial court noted that one of the two subpoenaed officers (not identified by name) had entered the courtroom. The identity was fixed subsequently at the federal evidentiary hearing, when Richard Montev-erde, the partner of Michael Stewart (the report’s author, who died prior to the hearing), testified that he was not the officer [439]*439referred to in the trial transcript. The record therefore supports the inference that the report’s author, Stewart, was present at trial. He did not testify.
Wilson’s trial counsel testified at the federal evidentiary hearing that he did not receive a copy of the report prior to or during trial, and did not learn of the report until the day before the evidentiary hearing. Wilson’s federal habeas counsel questioned trial counsel about the references to the report in the trial transcript:
Q. When you questioned the officers and the existence of a report was discussed, did you at that time know exactly what report that was?
A. No_ But it was obvious that there was some serious differences in the descriptions given beforehand and [Wilson] [sic ] at the time from what I had come to learn during the trial of the ease. And that’s all the questions are for[,] to attempt to determine from the police officer who handled the report[,] the prior descriptions.
It is apparent that, during trial, Wilson’s counsel was aware, at the very least, that a report existed, but was unaware of its contents. After learning at trial about the existence of the report, however, Wilson neither requested a copy nor asked the court to review it, in camera, to determine whether it contained any favorable evidence. And, although the trial judge, at the request of Wilson’s counsel, issued subpoenas for the investigating officers, and although Officer Stewart, the report’s author, was present at trial as a result and available to testify, Stewart was not called as a witness.
In light of these facts, we cannot conclude that the prosecutor’s failure to respond to Wilson’s general request for Brady material adversely affected trial counsel’s strategy. Accordingly, Wilson’s Brady request is not entitled to favorable weight in our assessment of the materiality of the undisclosed information. See Smith v. Black, 904 F.2d at 966 n. 4.
2.
In assessing the materiality of undisclosed impeachment evidence, “we must consider the nature of the impeachment evidence improperly withheld and the additional evidence of the defendant’s guilt independent of the disputed testimony”. United States v. Weintraub, 871 F.2d 1257, 1262 (5th Cir.1989).9 “The materiality of Brady material depends almost entirely on the value of the evidence relative to the other evidence mustered by the state”. Edmond v. Collins, 8 F.3d 290, 293 (5th Cir.1993) (internal quotation marks and citation omitted). For example, when the testimony of the witness who might have been impeached by the undisclosed evidence is strongly corroborated by additional evidence supporting a guilty verdict, the undisclosed evidence generally is not found to be material, Weintraub, 871 F.2d at 1262; but, on the other hand, “where the withheld evidence would seriously undermine the testimony of a key witness on an essential issue or there is no strong corroboration, the withheld evidence has been found to be material”.10 Id.
Nearly all of the evidence at trial consisted of Pierce’s identification testimony, supported by the testimony of Bowie, who was able to identify Wilson only by his build.11 There was no other corroborating evidence of Wilson’s guilt. Thus, Pierce’s eyewitness testimony was essential to Wilson’s conviction. Our court has noted that “[i]t is a commonplace that eyewitness testimony is highly regarded by juries, rather more than its objective appraisal might warrant”. Smith v. Black, 904 F.2d at 967. According[440]*440ly, in determining whether there is a reasonable probability that the outcome of the trial would have been different, our focus is on whether the report contains information that could have been used, on cross-examination, to significantly undermine Pierce’s credibility.12
As noted, of critical importance to our analysis of the materiality of any discrepancies is the manner by which the report was prepared. Monteverde, one of the officers who interviewed Pierce within hours of the robbery, testified at the federal evidentiary hearing that the report was prepared by his partner, Stewart, who died during the year prior to the hearing; that he had no independent recollection of the events reflected in the report; but that “[t]he normal procedure is to interview the victim or witness, ... write down the key points and then very soon thereafter relocate to an area where you can reflect and write the report, the narra-five”.13 Monteverde testified that the report is supposed to accurately reflect the victim’s statements to the officers. He testified further, however, that it is not a verbatim transcription of the victim’s statements.14
The first discrepancy between the report and Pierce’s testimony involves the assailants’ approach. The report states (Pierce is reported as having said) that two men came from behind Pierce and a third from behind the pumping station. Pierce testified that he saw one man cut across a field on his left, another came from behind the pumping station, and another came from behind him with a knife.15 But, the report contains, in addition to the narrative in which Pierce’s description of the assailants’ approach is reported, a description of the facial features of two of them. If Pierce had not seen their faces, he could not have described their facial features in the manner reflected in the report.16 In addition, it contains a section enti-[441]*441tied “Suspects Actions Before Offense”, in which boxes are checked, for both “wanted subject” no. 1 and no. 2, for “loiters in area”, “walks up to victim”, “approaches from behind”, and “follows victim on foot”.
The other discrepancies found to be material by the district court concern the identity of the perpetrator who struck Pierce in the face, pushed him down, and took his money. The report, quoted supra, states that “the first unk NM, with knife struck L. Pierce in face and struggle [ejnsued”; that Pierce was forced to the ground by “NM # 1”, who jabbed Pierce in the lower back and buttock, telling him to stay on the ground and give up the money; and that “Wanted Subject # 1” then “went into” Pierce’s back pocket, removing his wallet, and taking a $5.00 bill. At trial, Pierce testified that the perpetrator later identified as Wilson came up to him and tried to go through his pockets; that he hit Wilson and broke free, but then Wilson hit him, knocking his glasses off; that Wilson grabbed him from behind and wrestled him to the ground; that the “guy” with the knife placed the knife in his side; and then “they” went through his pockets and took his money-
The district court interpreted the report as using the terms “first unk NM”, “NM # 1”, and “Wanted Subject # 1” to refer to the same individual — an assailant other than Wilson. It therefore concluded that, according to the report, an assailant other than Wilson struck Pierce, wrestled him to the ground, and took his money — a version that differed materially from Pierce’s trial testimony that Wilson struck him and knocked his glasses off, grabbed him from behind and wrestled him to the ground, and that “they” took his money.
The district court’s interpretation of the cryptic labels used in the report to describe the perpetrators is certainly plausible. (One wonders why the author of the report used, in three consecutive sentences, three different terms (“first unk NM”, “NM # 1”, and “Wanted Subject # 1”) to describe, apparently, the same individual.) In any event, considering the circumstances of the report’s preparation, the fact that Pierce neither signed nor otherwise adopted the report, and the lack of any testimony about the report by either Pierce or its author, it is extremely difficult to evaluate the impact, if any, that disclosure of the report, and cross-examination of Pierce about the discrepancies between it and his testimony on direct examination, would have had on the outcome of the trial.
As our court stated in Lindsey v. King, 769 F.2d 1034, 1043 (5th Cir.1985), “[wjhether it is reasonably probable that a different result might have obtained had the evidence been disclosed [can be] a question of agonizing closeness”. In making the close call presented here, we must not focus solely on the discrepancies between Pierce’s testimony and the report. We must consider also items which are consistent: (1) the date, time, and location of the robbery, the number of assailants involved, and the amount of money taken from each of the victims are the same in both the report and Pierce’s testimony; (2) Wilson fits the physical description of “wanted subject #2” in the report, and Pierce’s description of Wilson at trial is consistent with the report’s description of Wilson;17 (3) Pierce’s testimony that the third suspect stood watch during the robbery is consistent with the report; and (4) Pierce’s testimony that Wilson held him while the other assailant stabbed him is consistent with the report, [442]*442which states that Pierce noticed a small incision made by “wanted subject # 1”. In addition, Pierce’s testimony regarding the robbery is generally consistent with the version of the robbery testified to by Bowie, the other victim.18
Finally, Wilson did not present a strong case for mistaken identity. Although 12 alibi witnesses testified that, at the time of the robbery, Wilson was playing football some distance away, we agree with the state appellate court’s characterization of their testimony as “less than definite”. State v. Wilson, 463 So.2d at 657. As that court noted, those witnesses “had no reason to fix events of the day of the robbery in their minds until weeks thereafter when [Wilson] was charged or, in most eases, until five or six months later when [Wilson’s] mother sought them out as witnesses”. Id. at 656.19
Because the report’s description of the robbery is subject to an interpretation that is less incriminating than that presented by Pierce’s testimony, the State should have disclosed it. But, although the conduct of the trial might have been affected by the failure to do so, we cannot conclude that there is a reasonable probability that, had it been disclosed, the outcome of the trial would have been different. Considering the incul-patory, not exculpatory, nature of the version in the report, the report’s consistency with much of Pierce’s testimony, Pierce’s opportunity to see Wilson during the robbery and his very definite identification of Wilson at trial, the consistency of Pierce’s and Bowie’s testimony, and the less than definite testimony of Wilson’s alibi witnesses, we cannot say that our confidence in the outcome of the trial has been undermined by the State’s failure to disclose the report.20
[443]*443At bottom, the Brady rule is one of the methods for seeking to ensure due process— a fair trial. Wilson received that. Accordingly, we REVERSE the judgment of the district court and REMAND for consideration of Wilson’s Sixth Amendment claim.
REVERSED AND REMANDED.