This case presents the issue as to whether the Due Process Clause of the Fourteenth Amendment compels the exclusion, in a state criminal trial, apart from any consideration of reliability, of pretrial identification evidence obtained by a police procedure that was both suggestive and unnecessary. This Court's decisions in Stovall v. Denno, 388 U.S. 293, 87 S.Ct. 1967, 18 L.Ed.2d 1199 (1967), and Neil v. Biggers, 409 U.S. 188, 93 S.Ct. 375, 34 L.Ed.2d 401 (1972), are particularly implicated.

* Jimmy D. Glover, a full-time trooper of the Connecticut State Police, in 1970 was assigned to the Narcotics Division in an undercover capacity. On May 5 of that year, about 7:45 p. m., e.d.t., and while there was still daylight, Glover and Henry Alton Brown, an informant, went to an apartment building at 201 Westland, in Hartford, for the purpose of purchasing narcotics from "Dickie Boy" Cicero, a known narcotics dealer. Cicero, it was thought, lived on the third floor of that apartment building. Tr. 45-46, 68.1 Glover and Brown entered the building, observed by back-up Officers D'Onofrio and Gaffey, and proceeded by stairs to the third floor. Glover knocked at the door of one of the two apartments served by the stairway.2 The area was illuminated by natural light from a window in the third floor hallway. Id., at 27-28. The door was opened 12 to 18 inches in response to the knock. Glover observed a man standing at the door and, behind him, a woman. Brown identified himself. Glover then asked for "two things" of narcotics. Id., at 29. The man at the door held out his hand, and Glover gave him two $10 bills. The door closed. Soon the man returned and handed Glover two glassine bags.3 While the door was open, Glover stood within two feet of the person from whom he made the purchase and observed his face. Five to seven minutes elapsed from the time the door first opened until it closed the second time. Id., at 30-33.

Glover and Brown then left the building. This was about eight minutes after their arrival. Glover drove to headquarters where he described the seller to D'Onofrio and Gaffey. Glover at that time did not know the identity of the seller. Id., at 36. He described him as being "a colored man, approximately five feet eleven inches tall, dark complexion, black hair, short Afro style, and having high cheekbones, and of heavy build. He was wearing at the time blue pants and a plaid shirt." Id., at 36-37. D'Onofrio, suspecting from this description that respondent might be the seller, obtained a photograph of respondent from the Records Division of the Hartford Police Department. He left it at Glover's office. D'Onofrio was not acquainted with respondent personally but did know him by sight and had seen him "(s)everal times" prior to May 5. Id., at 63-65. Glover, when alone, viewed the photograph for the first time upon his return to headquarters on May 7; he identified the person shown as the one from whom he had purchased the narcotics. Id., at 36-38.

The toxicological report on the contents of the glassine bags revealed the presence of heroin. The report was dated July 16, 1970. Id., at 75-76.

Respondent was arrested on July 27 while visiting at the apartment of a Mrs. Ramsey on the third floor of 201 Westland. This was the apartment at which the narcotics sale had taken place on May 5.4

Respondent was charged, in a two-count information, with possession and sale of heroin, in violation of Conn.Gen.Stat. (Rev. of 1958, as amended in 1969), §§ 19-481a and 19-480a (1977).5 At his trial in January 1971, the photograph from which Glover had identified respondent was received in evidence without objection on the part of the defense. Tr. 38. Glover also testified that, although he had not seen respondent in the eight months that had elapsed since the sale, "there (was) no doubt whatsoever" in his mind that the person shown on the photograph was respondent. Id., at 41-42. Glover also made a positive in-court identification without objection. Id., at 37-38.

No explanation was offered by the prosecution for the failure to utilize a photographic array or to conduct a lineup.

Respondent, who took the stand in his own defense, testified that on May 5, the day in question, he had been ill at his Albany Avenue apartment ("a lot of back pains, muscle spasms . . . a bad heart . . . high blood pressure . . . neuralgia in my face, and sinus," id., at 106), and that at no time on that particular day had he been at 201 Westland. Id., at 106, 113-114. His wife testified that she recalled, after her husband had refreshed her memory, that he was home all day on May 5. Id., at 164-165. Doctor Wesley M. Vietzke, an internist and assistant professor of medicine at the University of Connecticut, testified that respondent had consulted him on April 15, 1970, and that he took a medical history from him, heard his complaints about his back and facial pain, and discovered that he had high blood pressure. Id., at 129-131. The physician found respondent, subjectively, "in great discomfort." Id., at 135. Respondent in fact underwent surgery for a herniated disc at L5 and S1 on August 17. Id., at 157.

The jury found respondent guilty on both counts of the information. He received a sentence of not less than six nor more than nine years. His conviction was affirmed per curiam by the Supreme Court of Connecticut. State v. Brathwaite, 164 Conn. 617, 325 A.2d 284 (1973). That court noted the absence of an objection to Glover's in-court identification and concluded that respondent "has not shown that substantial injustice resulted from the admission of this evidence." Id., at 619, 325 A.2d, at 285. Under Connecticut law, substantial injustice must be shown before a claim of error not made or passed on by the trial court will be considered on appeal. Ibid.

Fourteen months later, respondent filed a petition for habeas corpus in the United States District Court for the District of Connecticut. He alleged that the admission of the identification testimony at his state trial deprived him of due process of law to which he was entitled under the Fourteenth Amendment. The District Court, by an unreported written opinion based on the court's review of the state trial transcript,6 dismissed respondent's petition. On appeal, the United States Court of Appeals for the Second Circuit reversed, with instructions to issue the writ unless the State gave notice of a desire to retry respondent and the new trial occurred within a reasonable time to be fixed by the District Judge.7 527 F.2d 363 (1975).

In brief summary, the court felt that evidence as to the photograph should have been excluded, regardless of reliability, because the examination of the single photograph was unnecessary and suggestive. And, in the court's view, the evidence was unreliable in any event. We granted certiorari. 425 U.S. 957, 96 S.Ct. 1737, 48 L.Ed.2d 202 (1976).

Stovall v. Denno, supra, decided in 1967, concerned a petitioner who had been convicted in a New York court of murder. He was arrested the day following the crime and was taken by the police to a hospital where the victim's wife, also wounded in the assault, was a patient. After observing Stovall and hearing him speak, she identified him as the murderer. She later made an in-court identification. On federal habeas, Stovall claimed the identification testimony violated his Fifth, Sixth, and Fourteenth Amendment rights. The District Court dismissed the petition, and the Court of Appeals, en banc, affirmed. This Court also affirmed. On the identification issue, the Court reviewed the practice of showing a suspect singly for purposes of identification, and the claim that this was so unnecessarily suggestive and conducive to irreparable mistaken identification that it constituted a denial of due process of law. The Court noted that the practice "has been widely condemned," 388 U.S., at 302, 87 S.Ct., at 1972, but it concluded that "a claimed violation of due process of law in the conduct of a confrontation depends on the totality of the circumstances surrounding it." Ibid. In that case, showing Stovall to the victim's spouse "was imperative." The Court then quoted the observations of the Court of Appeals, 2 Cir., 355 F.2d 731, 735 (CA2 1966), to the effect that the spouse was the only person who could possibly exonerate the accused; that the hospital was not far from the courthouse and jail; that no one knew how long she might live; that she was not able to visit the jail; and that taking Stovall to the hospital room was the only feasible procedure, and, under the circumstances, " 'the usual police station line-up . . . was out of the question.' " 388 U.S., at 302, 87 S.Ct., at 1972-1973.

Neil v. Biggers, supra, decided in 1972, concerned a respondent who had been convicted in a Tennessee court of rape, on evidence consisting in part of the victim's visual and voice identification of Biggers at a station-house showup seven months after the crime. The victim had been in her assailant's presence for some time and had directly observed him indoors and under a full moon outdoors. She testified that she had "no doubt" that Biggers was her assailant. She previously had given the police a description of the assailant. She had made no identification of others presented at previous showups, lineups, or through photographs. On federal habeas, the District Court held that the confrontation was so suggestive as to violate due process. The Court of Appeals affirmed. This Court reversed on that issue, and held that the evidence properly had been allowed to go to the jury. The Court reviewed Stovall and certain later cases where it had considered the scope of due process protection against the admission of evidence derived from suggestive identification procedures, namely, Simmons v. United States, 390 U.S. 377, 88 S.Ct. 967, 19 L.Ed.2d 1247 (1968); Foster v. California, 394 U.S. 440, 89 S.Ct. 1127, 22 L.Ed.2d 402 (1969); and Coleman v. Alabama, 399 U.S. 1, 90 S.Ct. 1999, 26 L.Ed.2d 387 (1970).8 The Court concluded that general guidelines emerged from these cases "as to the relationship between suggestiveness and misidentification." The "admission of evidence of a showup without more does not violate due process." 409 U.S., at 198, 93 S.Ct., at 382. The Court expressed concern about the lapse of seven months between the crime and the confrontation and observed that this "would be a seriously negative factor in most cases." Id., at 201, 93 S.Ct., at 383. The "central question," however, was "whether under the 'totality of the circumstances' the identification was reliable even though the confrontation procedure was suggestive." Id., at 199, 93 S.Ct., at 382. Applying that test, the Court found "no substantial likelihood of misidentification. The evidence was properly allowed to go to the jury." Id., at 201, 93 S.Ct., at 383.

Biggers well might be seen to provide an unambiguous answer to the question before us: The admission of testimony concerning a suggestive and unnecessary identification procedure does not violate due process so long as the identification possesses sufficient aspects of reliability.9 In one passage, however, the Court observed that the challenged procedure occurred pre-Stovall and that a strict rule would make little sense with regard to a confrontation that preceded the Court's first indication that a suggestive procedure might lead to the exclusion of evidence. Id., at 199, 93 S.Ct., at 382. One perhaps might argue that, by implication, the Court suggested that a different rule could apply post-Stovall. The question before us, then, is simply whether the Biggers analysis applies to post-Stovall confrontations as well to those pre-Stovall.

In the present case the District Court observed that the "sole evidence tying Brathwaite to the possession and sale of the heroin consisted in his identifications by the police undercover agent, Jimmy Glover." App. to Pet. for Cert. 6a. On the constitutional issue, the court stated that the first inquiry was whether the police used an impermissibly suggestive procedure in obtaining the out-of-court identification. If so, the second inquiry is whether, under all the circumstances, that suggestive procedure gave rise to a substantial likelihood of irreparable misidentification. Id., at 9a. Biggers and Simmons were cited. The court noted that in the Second Circuit, its controlling court, it was clear that "this type of identification procedure (display of a single photograph) is impermissibly suggestive," and turned to the second inquiry. App. to Pet. for Cert. 9a. The factors Biggers specified for consideration were recited and applied. The court concluded that there was no substantial likelihood of irreparable misidentification. It referred to the facts: Glover was within two feet of the seller. The duration of the confrontation was at least a "couple of minutes." There was natural light from a window or skylight and there was adequate light to see clearly in the hall. Glover "certainly was paying attention to identify the seller." Id., at 10a. He was a trained police officer who realized that later he would have to find and arrest the person with whom he was dealing. He gave a detailed description to D'Onofrio. The reliability of this description was supported by the fact that it enabled D'Onofrio to pick out a single photograph that was thereafter positively identified by Glover. Only two days elapsed between the crime and the photographic identification. Despite the fact that another eight months passed before the in-court identification, Glover had "no doubt" that Brathwaite was the person who had sold him heroin.

The Court of Appeals confirmed that the exhibition of the single photograph to Glover was "impermissibly suggestive," 527 F.2d, at 366, and felt that, in addition, "it was unnecessarily so." Id., at 367. There was no emergency and little urgency. The court said that prior to the decision in Biggers, except in cases of harmless error, "a conviction secured as the result of admitting an identification obtained by impermissibly suggestive and unnecessary measures could not stand." Ibid. It noted what it felt might be opposing inferences to be drawn from passages in Biggers, but concluded that the case preserved the principle "requiring the exclusion of identifications resulting from 'unnecessarily suggestive confrontation' " in post-Stovall situations. 527 F.2d, at 368. The court also concluded that for post-Stovall identifications, Biggers had not changed the existing rule. Thus: "Evidence of an identification unnecessarily obtained by impermissibly suggestive means must be excluded under Stovall . . . . No rules less stringent than these can force police administrators and prosecutors to adopt procedures that will give fair assurance against the awful risks of misidentification." 527 F.2d, at 371. Finally, the court said, even if this conclusion were wrong, the writ, nevertheless, should issue. It took judicial notice that on May 5, 1970, sunset at Hartford was at 7:53 p. m. It characterized Glover's duty as an undercover agent as one "to cause arrests to be made," and his description of the suspect as one that "could have applied to hundreds of Hartford black males." Ibid. The in-court identification had "little meaning," for Brathwaite was at the counsel table. The fact that respondent was arrested in the very apartment where the sale was made was subject to a "not implausible" explanation from the respondent, "although evidently not credited by the jury." And the court was troubled by "the long and unexplained delay" in the arrest. It was too great a danger that the respondent was convicted because he was a man D'Onofrio had previously observed near the scene, was thought to be a likely offender, and was arrested when he was known to be in Mrs. Ramsey's apartment, rather than because Glover "really remembered him as the seller." Id., at 371-372.

Petitioner at the outset acknowledges that "the procedure in the instant case was suggestive (because only one photograph was used) and unnecessary" (because there was no emergency or exigent circumstance). Brief for Petitioner 10; Tr. of Oral Arg. 7. The respondent, in agreement with the Court of Appeals, proposes a per se rule of exclusion that he claims is dictated by the demands of the Fourteenth Amendment's guarantee of due process. He rightly observes that this is the first case in which this Court has had occasion to rule upon strictly post-Stovall out-of-court identification evidence of the challenged kind.

Since the decision in Biggers, the Courts of Appeals appear to have developed at least two approaches to such evidence. See Pulaski, Neil v. Biggers : The Supreme Court Dismantles the Wade Trilogy's Due Process Protection, 26 Stan.L.Rev. 1097, 1111-1114 (1974). The first, or per se approach, employed by the Second Circuit in the present case, focuses on the procedures employed and requires exclusion of the out-of-court identification evidence, without regard to reliability, whenever it has been obtained through unnecessarily suggested confrontation procedures.10 The justifications advanced are the elimination of evidence of uncertain reliability, deterrence of the police and prosecutors, and the stated "fair assurance against the awful risks of misidentification." 527 F.2d, at 371. See Smith v. Coiner, 473 F.2d 877, 882 (CA4), cert. denied sub nom. Wallace v. Smith, 414 U.S. 1115, 94 S.Ct. 848, 38 L.Ed.2d 743 (1973).

The second, or more lenient, approach is one that continues to rely on the totality of the circumstances. It permits the admission of the confrontation evidence if, despite the suggestive aspect, the out-of-court identification possesses certain features of reliability. Its adherents feel that the per se approach is not mandated by the Due Process Clause of the Fourteenth Amendment. This second approach, in contrast to the other, is ad hoc and serves to limit the societal costs imposed by a sanction that excludes relevant evidence from consideration and evaluation by the trier of fact. See United States ex rel. Kirby v. Sturges, 510 F.2d 397, 407-408 (CA7) (opinion by Judge, now Mr. Justice, Stevens), cert. denied, 421 U.S. 1016, 95 S.Ct. 2424, 44 L.Ed.2d 685 (1975); Stanley v. Cox, 486 F.2d 48 (CA4 1973), cert. denied sub nom. Stanley v. Slayton, 416 U.S. 958, 94 S.Ct. 1975, 40 L.Ed.2d 309 (1974).11

Mr. Justice Stevens, in writing for the Seventh Circuit in Kirby, supra, observed: "There is surprising unanimity among scholars in regarding such a rule (the per se approach) as essential to avoid serious risk of miscarriage of justice." 510 F.2d, at 405. He pointed out that well-known federal judges have taken the position that "evidence of, or derived from, a showup identification should be inadmissible unless the prosecutor can justify his failure to use a more reliable identification procedure." Id., at 406. Indeed, the ALI Model Code of Pre-Arraignment Procedure §§ 160.1 and 160.2 (1975) (hereafter Model Code), frowns upon the use of a showup or the display of only a single photograph.

The respondent here stresses the same theme and the need for deterrence of improper identification practice, a factor he regards as pre-eminent. Photographic identification, it is said, continues to be needlessly employed. He notes that the legislative regulation "the Court had hoped (United States v.) Wade (, 388 U.S. 218, 239, 87 S.Ct. 1926, 1938-1939, 18 L.Ed.2d 1149 (1967),) would engender," Brief for Respondent 15, has not been forthcoming. He argues that a totality rule cannot be expected to have a significant deterrent impact; only a strict rule of exclusion will have direct and immediate impact on law enforcement agents. Identification evidence is so convincing to the jury that sweeping exclusionary rules are required. Fairness of the trial is threatened by suggestive confrontation evidence, and thus, it is said, an exclusionary rule has an established constitutional predicate.

There are, of course, several interests to be considered and taken into account. The driving force behind United States v. Wade, 388 U.S. 218, 87 S.Ct. 1926, 18 L.Ed.2d 1149 (1967), Gilbert v. California, 388 U.S. 263, 87 S.Ct. 1951, 18 L.Ed.2d 1178 (1967) (right to counsel at a post-indictment line-up), and Stovall, all decided on the same day, was the Court's concern with the problems of eyewitness identification. Usually the witness must testify about an encounter with a total stranger under circumstances of emergency or emotional stress. The witness' recollection of the stranger can be distorted easily by the circumstances or by later actions of the police. Thus, Wade and its companion cases reflect the concern that the jury not hear eyewitness testimony unless that evidence has aspects of reliability. It must be observed that both approaches before us are responsive to this concern. The per se rule, however, goes too far since its application automatically and peremptorily, and without consideration of alleviating factors, keeps evidence from the jury that is reliable and relevant.

The second factor is deterrence. Although the per se approach has the more significant deterrent effect, the totality approach also has an influence on police behavior. The police will guard against unnecessarily suggestive procedures under the totality rule, as well as the per se one, for fear that their actions will lead to the exclusion of identifications as unreliable.12

The third factor is the effect on the administration of justice. Here the per se approach suffers serious drawbacks. Since it denies the trier reliable evidence, it may result, on occasion, in the guilty going free. Also, because of its rigidity, the per se approach may make error by the trial judge more likely than the totality approach. And in those cases in which the admission of identification evidence is error under the per se approach but not under the totality approach cases in which the identification is reliable despite an unnecessarily suggestive identification procedure reversal is a Draconian sanction.13 Certainly, inflexible rules of exclusion that may frustrate rather than promote justice have not been viewed recently by this Court with unlimited enthusiasm. See, for example, the several opinions in Brewer v. Williams, 430 U.S. 387, 97 S.Ct. 1232, 51 L.Ed.2d 424 (1977). See also United States v. Janis, 428 U.S. 433, 96 S.Ct. 3021, 49 L.Ed.2d 1046 (1976).

It is true, as has been noted, that the Court in Biggers referred to the pre-Stovall character of the confrontation in that case. 409 U.S., at 199, 93 S.Ct., at 382. But that observation was only one factor in the judgmental process. It does not translate into a holding that post-Stovall confrontation evidence automatically is to be excluded.

The standard, after all, is that of fairness as required by the Due Process Clause of the Fourteenth Amendment. See United States v. Lovasco, 431 U.S. 783, 790, 97 S.Ct. 2044, 2049, 52 L.Ed.2d 752 (1977); Rochin v. California, 342 U.S. 165, 170-172, 72 S.Ct. 205, 208-209, 96 L.Ed. 183 (1952). Stovall, with its reference to "the totality of the circumstances," 388 U.S., at 302, 87 S.Ct., at 1972, and Biggers, with its continuing stress on the same totality, 409 U.S., at 199, 93 S.Ct., at 382, did not, singly or together, establish a strict exclusionary rule or new standard of due process. Judge Leventhal, although speaking pre-Biggers and of a pre-Wade situation, correctly has described Stovall as protecting an evidentiary interest and, at the same time, as recognizing the limited extent of that interest in our adversary system.14 sWe therefore conclude that reliability is the linchpin in determining the admissibility of identification testimony for both pre- and post-Stovall confrontations. The factors to be considered are set out in Biggers. 409 U.S., at 199-200, 93 S.Ct., at 382. These include the opportunity of the witness to view the criminal at the time of the crime, the witness' degree of attention, the accuracy of his prior description of the criminal, the level of certainty demonstrated at the confrontation, and the time between the crime and the confrontation. Against these factors is to be weighed the corrupting effect of the suggestive identification itself.

We turn, then, to the facts of this case and apply the analysis:

1. The opportunity to view. Glover testified that for two to three minutes he stood at the apartment door, within two feet of the respondent. The door opened twice, and each time the man stood at the door. The moments passed, the conversation took place, and payment was made. Glover looked directly at his vendor. It was near sunset, to be sure, but the sun had not yet set, so it was not dark or even dusk or twilight. Natural light from outside entered the hallway through a window. There was natural light, as well, from inside the apartment.

2. The degree of attention. Glover was not a casual or passing observer, as is so often the case with eyewitness identification. Trooper Glover was a trained police officer on duty and specialized and dangerous duty when he called at the third floor of 201 Westland in Hartford on May 5, 1970. Glover himself was a Negro and unlikely to perceive only general features of "hundreds of Hartford black males," as the Court of Appeals stated. 527 F.2d, at 371. It is true that Glover's duty was that of ferreting out narcotics offenders and that he would be expected in his work to produce results. But it is also true that, as a specially trained, assigned, and experienced officer, he could be expected to pay scrupulous attention to detail, for he knew that subsequently he would have to find and arrest his vendor. In addition, he knew that his claimed observations would be subject later to close scrutiny and examination at any trial.

3. The accuracy of the description. Glover's description was given to D'Onofrio within minutes after the transaction. It included the vendor's race, his height, his build, the color and style of his hair, and the high cheekbone facial feature. It also included clothing the vendor wore. No claim has been made that respondent did not possess the physical characteristics so described. D'Onofrio reacted positively at once. Two days later, when Glover was alone, he viewed the photograph D'Onofrio produced and identified its subject as the narcotics seller.

4. The witness' level of certainty. There is no dispute that the photograph in question was that of respondent. Glover, in response to a question whether the photograph was that of the person from whom he made the purchase, testified: "There is no question whatsoever." Tr. 38. This positive assurance was repeated. Id., at 41-42.

5. The time between the crime and the confrontation. Glover's description of his vendor was given to D'Onofrio within minutes of the crime. The photographic identification took place only two days later. We do not have here the passage of weeks or months between the crime and the viewing of the photograph.

These indicators of Glover's ability to make an accurate identification are hardly outweighed by the corrupting effect of the challenged identification itself. Although identifications arising from single-photograph displays may be viewed in general with suspicion, see Simmons v. United States, 390 U.S., at 383, 88 S.Ct., at 970-971, we find in the instant case little pressure on the witness to acquiesce in the suggestion that such a display entails. D'Onofrio had left the photograph at Glover's office and was not present when Glover first viewed it two days after the event. There thus was little urgency and Glover could view the photograph at his leisure. And since Glover examined the photograph alone, there was no coercive pressure to make an identification arising from the presence of another. The identification was made in circumstances allowing care and reflection.

Although it plays no part in our analysis, all this assurance as to the reliability of the identification is hardly undermined by the facts that respondent was arrested in the very apartment where the sale had taken place, and that he acknowledged his frequent visits to that apartment.15

Surely, we cannot say that under all the circumstances of this case there is "a very substantial likelihood of irreparable misidentification." Id., at 384, 88 S.Ct., at 971. Short of that point, such evidence is for the jury to weigh. We are content to rely upon the good sense and judgment of American juries, for evidence with some element of untrustworthiness is customary grist for the jury mill. Juries are not so susceptible that they cannot measure intelligently the weight of identification testimony that has some questionable feature.

Of course, it would have been better had D'Onofrio presented Glover with a photographic array including "so far as practicable . . . a reasonable number of persons similar to any person then suspected whose likeness is included in the array." Model Code, § 160.2(2). The use of that procedure would have enhanced the force of the identification at trial and would have avoided the risk that the evidence would be excluded as unreliable. But we are not disposed to view D'Onofrio's failure as one of constitutional dimension to be enforced by a rigorous and unbending exclusionary rule. The defect, if there be one, goes to weight and not to substance.16

We conclude that the criteria laid down in Biggers are to be applied in determining the admissibility of evidence offered by the prosecution concerning a post-Stovall identification, and that those criteria are satisfactorily met and complied with here.

The judgment of the Court of Appeals is reversed.

It is so ordered.

Mr. Justice STEVENS, concurring.

While I join the Court's opinion, I would emphasize two points.

First, as I indicated in my opinion in United States ex rel. Kirby v. Sturges, 510 F.2d 397, 405-406 (CA7 1975), the arguments in favor of fashioning new rules to minimize the danger of convicting the innocent on the basis of unreliable eyewitness testimony carry substantial force. Nevertheless, for the reasons stated in that opinion, as well as those stated by the Court today. I am persuaded that this rulemaking function can be performed "more effectively by the legislative process than by a somewhat clumsy judicial fiat," id., at 408, and that the Federal Constitution does not foreclose experimentation by the States in the development of such rules.

Second, in evaluating the admissibility of particular identification testimony it is sometimes difficult to put other evidence of guilt entirely to one side.* Mr. Justice BLACKMUN'S opinion for the Court carefully avoids this pitfall and correctly relies only on appropriate indicia of the reliability of the identification itself. Although I consider the factual question in this case extremely close, I am persuaded that the Court has resolved it properly.

Mr. Justice MARSHALL, with whom Mr. Justice BRENNAN joins, dissenting.

Today's decision can come as no surprise to those who have been watching the Court dismantle the protections against mistaken eyewitness testimony erected a decade ago in United States v. Wade, 388 U.S. 218, 87 S.Ct. 1926, 18 L.Ed.2d 1149 (1967); Gilbert v. California, 388 U.S. 263, 87 S.Ct. 1951, 18 L.Ed.2d 1178 (1967); and Stovall v. Denno, 388 U.S. 293, 87 S.Ct. 1967, 18 L.Ed.2d 1199 (1967). But it is still distressing to see the Court virtually ignore the teaching of experience embodied in those decisions and blindly uphold the conviction of a defendant who may well be innocent.

* The magnitude of the Court's error can be seen by analyzing the cases in the Wade trilogy and the decisions following it. The foundation of the Wade trilogy was the Court's recognition of the "high incidence of miscarriage of justice" resulting from the admission of mistaken eyewitness identification evidence at criminal trials. United States v. Wade, supra, at 228, 87 S.Ct., at 1933. Relying on numerous studies made over many years by such scholars as Professor Wigmore and Mr. Justice Frankfurter, the Court concluded that "(t)he vagaries of eyewitness identification are well-known; the annals of criminal law are rife with instances of mistaken identification." Ibid. It is, of course, impossible to control one source of such errors the faulty perceptions and unreliable memories of witnesses except through vigorously contested trials conducted by diligent counsel and judges. The Court in the Wade cases acted, however, to minimize the more preventable threat posed to accurate identification by "the degree of suggestion inherent in the manner in which the prosecution presents the suspect to witnesses for pretrial identification." Ibid.

The Court did so in Wade and Gilbert v. California by prohibiting the admission at trial of evidence of pretrial confrontations at which an accused was not represented by counsel. Further protection was afforded by holding that an in-court identification following an uncounseled lineup was allowable only if the prosecution could clearly and convincingly demonstrate that it was not tainted by the constitutional violation. Only in this way, the Court held, could confrontations fraught with the danger of misidentification be made fairer, and could Sixth Amendment rights to assistance of counsel and confrontation of witnesses at trial be effectively preserved. The crux of the Wade decisions, however, was the unusual threat to the truth-seeking process posed by the frequent untrustworthiness of eyewitness identification testimony. This, combined with the fact that juries unfortunately are often unduly receptive to such evidence,1 is the fundamental fact of judicial experience ignored by the Court today.

Stovall v. Denno, while holding that the Wade prophylactic rules were not retroactive, was decided at the same time and reflects the same concerns about the reliability of identification testimony. Stovall recognized that, regardless of Sixth Amendment principles, "the conduct of a confrontation" may be "so unnecessarily suggestive and conducive to irreparable mistaken identification" as to deny due process of law. 388 U.S., at 301-302, 87 S.Ct., at 1972. The pretrial confrontation in Stovall was plainly suggestive,2 and evidence of it was introduced at trial along with the witness' in-court identification. The Court ruled that there had been no violation of due process, however, because the unusual necessity for the procedure3 outweighed the danger of suggestion.

Stovall thus established a due process right of criminal suspects to be free from confrontations that, under all the circumstances, are unnecessarily suggestive. The right was enforceable by exclusion at trial of evidence of the constitutionally invalid identification. Comparison with Wade and Gilbert confirms this interpretation. Where their Sixth Amendment holding did not apply, Stovall found an analogous Fourteenth Amendment right to a lineup conducted in a fundamentally fair manner. This interpretation is reinforced by the Court's statement that "a claimed violation of due process of law in the conduct of a confrontation depends on the totality of the circumstances surrounding it." 388 U.S., at 302, 87 S.Ct., at 1972 (emphasis added). Significantly, several years later, Stovall was viewed in precisely the same way, even as the Court limited Wade and Gilbert to post-indictment confrontations: "The Due Process Clause . . . forbids a lineup that is unnecessarily suggestive and conducive to irreparable mistaken identification. Stovall v. Denno, 388 U.S. 293, 87 S.Ct. 1967, 18 L.Ed.2d 1199; Foster v. California, 394 U.S. 440, 89 S.Ct. 1127, 22 L.Ed.2d 402." Kirby v. Illinois, 406 U.S. 682, 691, 92 S.Ct. 1877, 1883, 32 L.Ed.2d 411 (1972) (emphasis added).4

The development of due process protections against mistaken identification evidence, begun in Stovall, was continued in Simmons v. United States, 390 U.S. 377, 88 S.Ct. 967, 19 L.Ed.2d 1247 (1968). There, the Court developed a different rule to deal with the admission of in-court identification testimony that the accused claimed had been fatally tainted by a previous suggestive confrontation. In Simmons, the exclusionary effect of Stovall had already been accomplished, since the prosecution made no use of the suggestive confrontation. Simmons, therefore, did not deal with the constitutionality of the pretrial identification procedure. The only question was the impact of the Due Process Clause on an in-court identification that was not itself unnecessarily suggestive. Simmons held that due process was violated by the later identification if the pretrial procedure had been "so impermissibly suggestive as to give rise to a very substantial likelihood of irreparable misidentification." 390 U.S., at 384, 88 S.Ct., at 971. This test focused, not on the necessity for the challenged pretrial procedure, but on the degree of suggestiveness that it entailed. In applying this test, the Court understandably considered the circumstances surrounding the witnesses' initial opportunity to view the crime. Finding that any suggestion in the pretrial confrontation had not affected the fairness of the in-court identification, Simmons rejected petitioner's due process attack on his conviction.

Again, comparison with the Wade cases is instructive. The inquiry mandated by Simmons is similar to the independent-source test used in Wade where an in-court identification is sought following an uncounseled lineup. In both cases, the issue is whether the witness is identifying the defendant solely on the basis of his memory of events at the time of the crime, or whether he is merely remembering the person he picked out in a pretrial procedure. Accordingly, in both situations, the relevant inquiry includes factors bearing on the accuracy of the witness' identification, including his opportunity to view the crime.

Thus, Stovall and Simmons established two different due process tests for two very different situations. Where the prosecution sought to use evidence of a questionable pretrial identification, Stovall required its exclusion, because due process had been violated by the confrontation, unless the necessity for the unduly suggestive procedure outweighed its potential for generating an irreparably mistaken identification. The Simmons test, on the other hand, was directed to ascertaining due process violations in the introduction of in-court identification testimony that the defendant claimed was tainted by pretrial procedures. In the latter situation, a court could consider the reliability of the identification under all the circumstances.5

This distinction between Stovall and Simmons was preserved in two succeeding cases. Foster v. California, 394 U.S. 440, 89 S.Ct. 1127, 22 L.Ed.2d 402 (1969), like Stovall, involved both unduly suggestive pretrial procedures, evidence of which was introduced at trial, and a tainted in-court identification. Accordingly, Foster applied the Stovall test, 394 U.S., at 442, 89 S.Ct., at 1128, and held that the police "procedure so undermined the reliability of the eyewitness identification as to violate due process." Id., at 443, 89 S.Ct., at 1129 (emphasis added). In contrast, in Coleman v. Alabama, 399 U.S. 1, 90 S.Ct. 1999, 26 L.Ed.2d 387 (1970), where the witness' pretrial identification was not used to bolster his in-court identification, the plurality opinion applied the test enunciated in Simmons. It concluded that an in-court identification did not violate due process because it did not stem from an allegedly suggestive lineup.

The Court inexplicably seemed to erase the distinction between Stovall and Simmons situations in Neil v. Biggers, 409 U.S. 188, 93 S.Ct. 375, 34 L.Ed.2d 401 (1972). In Biggers there was a pretrial confrontation that was clearly both suggestive and unnecessary.6 Evidence of this, together with an in-court identification, was admitted at trial. Biggers was, in short, a case plainly cast in the Stovall mold. Yet the Court, without explanation or apparent recognition of the distinction, applied the Simmons test. The Court stated: "(T)he primary evil to be avoided is 'a very substantial likelihood of irreparable misidentification.' Simmons v. United States, 390 U.S., at 384, 88 S.Ct., at 971. . . . It is the likelihood of misidentification which violates a defendant's right to due process. . . ." 409 U.S., at 198, 93 S.Ct., at 381-382. While this statement accurately describes the lesson of Simmons, it plainly ignores the teaching of Stovall and Foster that an unnecessarily suggestive pretrial confrontation itself violates due process.

But the Court did not simply disregard the due process analysis of Stovall. It went on to take the Simmons standard for assessing the constitutionality of an in-court identification " 'a very substantial likelihood of irreparable misidentification' " and transform it into the "standard for the admissibility of testimony concerning (an) out-of-court identification." 409 U.S., at 198, 93 S.Ct., at 381. It did so by deleting the word "irreparable" from the Simmons formulation. This metamorphosis could be accomplished, however, only by ignoring the fact that Stovall fortified only months earlier by Kirby v. Illinois, see supra, at 121, had established a test for precisely the same situation that focused on the need for the suggestive procedure. It is not surprising that commentators almost unanimously mourned the demise of Stovall in the Biggers decision.7

Apparently, the Court does not consider Biggers controlling in this case. I entirely agree, since I believe that Biggers was wrongly decided. The Court, however, concludes that Biggers is distinguishable because it, like the identification decisions that preceded it, involved a pre-Stovall confrontation, and because a paragraph in Biggers itself, 409 U.S., at 198-199, 93 S.Ct., at 381-382, seems to distinguish between pre- and post-Stovall confrontations. Accordingly, in determining the admissibility of the post-Stovall identification in this case, the Court considers two alternatives, a per se exclusionary rule and a totality-of-the circumstances approach. Ante, at 110-111. The Court weighs three factors in deciding that the totality approach, which is essentially the test used in Biggers, should be applied. Ante, at 111-113. In my view, the Court wrongly evaluates the impact of these factors.

First, the Court acknowledges that one of the factors, deterrence of police use of unnecessarily suggestive identification procedures, favors the per se rule. Indeed, it does so heavily, for such a rule would make it unquestionably clear to the police they must never use a suggestive procedure when a fairer alternative is available. I have no doubt that conduct would quickly conform to the rule.

Second, the Court gives passing consideration to the dangers of eyewitness identification recognized in the Wade trilogy. It concludes, however, that the grave risk of error does not justify adoption of the per se approach because that would too often result in exclusion of relevant evidence. In my view, this conclusion totally ignores the lessons of Wade. The dangers of mistaken identification are, as Stovall held, simply too great to permit unnecessarily suggestive identifications. Neither Biggers nor the Court's opinion today points to any contrary empirical evidence. Studies since Wade have only reinforced the validity of its assessment of the dangers of identification testimony.8 While the Court is "content to rely on the good sense and judgment of American juries," ante, at 116, the impetus for Stovall and Wade was repeated miscarriages of justice resulting from juries' willingness to credit inaccurate eyewitness testimony.

Finally, the Court errs in its assessment of the relative impact of the two approaches on the administration of justice. The Court relies most heavily on this factor, finding that "reversal is a Draconian sanction" in cases where the identification is reliable despite an unnecessarily suggestive procedure used to obtain it. Relying on little more than a strong distaste for "inflexible rules of exclusion," the Court rejects the per se test. Ante, at 113. In so doing, the Court disregards two significant distinctions between the per se rule advocated in this case and the exclusionary remedies for certain other constitutional violations.

First, the per se rule here is not "inflexible." Where evidence is suppressed, for example, as the fruit of an unlawful search, it may well be forever lost to the prosecution. Identification evidence, however, can by its very nature be readily and effectively reproduced. The in-court identification, permitted under Wade and Simmons if it has a source independent of an uncounseled or suggestive procedure, is one example. Similarly, when a prosecuting attorney learns that there has been a suggestive confrontation, he can easily arrange another lineup conducted under scrupulously fair conditions. Since the same factors are evaluated in applying both the Court's totality test and the Wade-Simmons independent-source inquiry, any identification which is "reliable" under the Court's test will support admission of evidence concerning such a fairly conducted lineup. The evidence of an additional, properly conducted confrontation will be more persuasive to a jury, thereby increasing the chance of a justified conviction where a reliable identification was tainted by a suggestive confrontation. At the same time, however, the effect of an unnecessarily suggestive identification which has no value whatsoever in the law enforcement process will be completely eliminated.

Second, other exclusionary rules have been criticized for preventing jury consideration of relevant and usually reliable evidence in order to serve interest unrelated to guilt or innocence, such as discouraging illegal searches or denial of counsel. Suggestively obtained eyewitness testimony is excluded, in contrast, precisely because of its unreliability and concomitant irrelevance. Its exclusion both protects the integrity of the truth-seeking function of the trial and discourages police use of needlessly inaccurate and ineffective investigatory methods.

Indeed, impermissibly suggestive identifications are not merely worthless law enforcement tools. They pose a grave threat to society at large in a more direct way than most governmental disobedience of the law, see Olmstead v. United States, 277 U.S. 438, 471, 485, 48 S.Ct. 564, 570, 575, 72 L.Ed. 944 (1928) (Brandeis, J., dissenting). For if the police and the public erroneously conclude, on the basis of an unnecessarily suggestive confrontation, that the right man has been caught and convicted, the real outlaw must still remain at large. Law enforcement has failed in its primary function and has left society unprotected from the depredations of an active criminal.

For these reasons, I conclude that adoption of the per se rule would enhance, rather than detract from, the effective administration of justice. In my view, the Court's totality test will allow seriously unreliable and misleading evidence to be put before juries. Equally important, it will allow dangerous criminals to remain on the streets while citizens assume that police action has given them protection. According to my calculus, all three of the factors upon which the Court relies point to acceptance of the per se approach.

Even more disturbing than the Court's reliance on the totality test, however, is the analysis it uses, which suggests a reinterpretation of the concept of due process of law in criminal cases. The decision suggests that due process violations in identification procedures may not be measured by whether the government employed procedures violating standards of fundamental fairness. By relying on the probable accuracy of a challenged identification, instead of the necessity for its use, the Court seems to be ascertaining whether the defendant was probably guilty. Until today, I had though that "Equal justice under law" meant that the existence of constitutional violations did not depend on the race, sex, religion, nationality, or likely guilt of the accused. The Due Process Clause requires adherence to the same high standard of fundamental fairness in dealing with every criminal defendant, whatever his personal characteristics and irrespective of the strength of the State's case against him. Strong evidence that the defendant is guilty should be relevant only to the determination whether an error of constitutional magnitude was nevertheless harmless beyond a reasonable doubt. See Chapman v. California, 386 U.S. 18, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967). By importing the question of guilt into the initial determination of whether there was a constitutional violation, the apparent effect of the Court's decision is to undermine the protection afforded by the Due Process Clause. "It is therefore important to note that the state courts remain free, in interpreting state constitutions, to guard against the evil clearly identified by this case." Oregon v. Mathiason, 429 U.S. 492, 499, 97 S.Ct. 711, 716, 50 L.Ed.2d 714 (1977) (Marshall, J., dissenting).9

Despite my strong disagreement with the Court over the proper standards to be applied in this case, I am pleased that its application of the totality test does recognize the continuing vitality of Stovall. In assessing the reliability of the identification, the Court mandates weighing "the corrupting effect of the suggestive identification itself" against the "indicators of (a witness') ability to make an accurate identification." Ante, at 114, 116. The Court holds, as Neil v. Biggers failed to, that a due process identification inquiry must take account of the suggestiveness of a confrontation and the likelihood that it led to misidentification, as recognized in Stovall and Wade. Thus, even if a witness did have an otherwise adequate opportunity to view a criminal, the later use of a highly suggestive identification procedure can render his testimony inadmissible. Indeed, it is my view that, assuming applicability of the totality test enunciated by the Court, the facts of the present case require that result.

I consider first the opportunity that Officer Glover had to view the suspect. Careful review of the record shows that he could see the heroin seller only for the time it took to speak three sentences of four or five short words, to hand over some money, Tr. 29-30, and later after the door reopened, to receive the drugs in return, id., at 30, 31-32. The entire face-to-face transaction could have taken as little as 15 or 20 seconds. But during this time, Glover's attention was not focused exclusively on the seller's face. He observed that the door was opened 12 to 18 inches id., at 29, that there was a window in the room behind the door, id., at 33, and, most importantly, that there was a woman standing behind the man, id., at 29, 30. Glover was, of course, also concentrating on the details of the transaction he must have looked away from the seller's face to hand him the money and receive the drugs. The observation during the conversation thus may have been as brief as 5 or 10 seconds.

As the Court notes, Glover was a police officer trained in and attentive to the need for making accurate identifications. Nevertheless, both common sense and scholarly study indicate that while a trained observer such as a police officer "is somewhat less likely to make an erroneous identification than the average untrained observer, the mere fact that he has been so trained is no guarantee that he is correct in a specific case. His identification testimony should be scrutinized just as carefully as that of the normal witness." Wall, supra, n. 1, at 14; see also Levine & Tapp, supra, n. 8, at 1088. Moreover, "identifications made by policemen in highly competitive activities, such as undercover narcotic agents . . ., should be scrutinized with special care." Wall, supra, n. 1, at 14. Yet it is just such a searching inquiry that the Court fails to make here.

Another factor on which the Court relies the witness' degree of certainty in making the identification is worthless as an indicator that he is correct.10 Even if Glover had been unsure initially about his identification of respondent's picture, by the time he was called at trial to present a key piece of evidence for the State that paid his salary, it is impossible to imagine his responding negatively to such questions as "is there any doubt in your mind whatsoever" that the identification was correct. Tr. 34, 41-42. As the Court noted in Wade: " 'It is a matter of common experience that, once a witness has picked out the accused at the (pretrial confrontation), he is not likely to go back on his word later on.' " 388 U.S., at 229, 87 S.Ct., at 1933, quoting Williams & Hammelmann, Identification Parades I, Crim.L.Rev. 479, 482 (1963).

Next, the Court finds that because the identification procedure took place two days after the crime, its reliability is enhanced. While such temporal proximity makes the identification more reliable than one occurring months later, the fact is that the greatest memory loss occurs within hours after an event. After that, the dropoff continues much more slowly.11 Thus, the reliability of an identification is increased only if it was made within several hours of the crime. If the time gap is any greater, reliability necessarily decreases.

Finally, the Court makes much of the fact that Glover gave a description of the seller to D'Onofrio shortly after the incident. Despite the Court's assertion that because "Glover himself was a Negro and unlikely to perceive only general features of 'hundreds of Hartford black males,' as the Court of Appeals stated," ante, at 115, the description given by Glover was actually no more than a general summary of the seller's appearance. See, ante, at 101. We may discount entirely the seller's clothing, for that was of no significance later in the proceeding. Indeed, to the extent that Glover noticed clothes, his attention was diverted from the seller's face. Otherwise, Glover merely described vaguely the seller's height, skin color, hairstyle, and build. He did say that the seller had "high cheekbones," but there is no other mention of facial features, nor even an estimate of age. Conspicuously absent is any indication that the seller was a native of the West Indies, certainly something which a member of the black community could immediately recognize from both appearance and accent.12

From all of this, I must conclude that the evidence of Glover's ability to make an accurate identification is far weaker than the Court finds it. In contrast, the procedure used to identify respondent was both extraordinarily suggestive and strongly conducive to error. In dismissing "the corrupting effect of the suggestive identification" procedure here, ante, at 116, the Court virtually grants the police license to convict the innocent. By displaying a single photograph of respondent to the witness Glover under the circumstances in this record almost everything that could have been done wrong was done wrong.

In the first place, there was no need to use a photograph at all. Because photos are static, two-dimensional, and often outdated, they are "clearly inferior in reliability" to corporeal procedures. Wall, supra, n. 1, at 70; People v. Gould, 54 Cal.2d 621, 631, 7 Cal.Rptr. 273, 278, 354 P.2d 865, 870 (1960). While the use of photographs is justifiable and often essential where the police have no knowledge of an offender's identity, the poor reliability of photos makes their use inexcusable where any other means of identification is available. Here, since Detective D'Onofrio believed that he knew the seller's identity, see ante, at 101, 115, further investigation without resort to a photographic showup was easily possible. With little inconvenience, a corporeal lineup including Brathwaite might have been arranged.13 Properly conducted, such a procedure would have gone far to remove any doubt about the fairness and accuracy of the identification.14

Worse still than the failure to use an easily available corporeal identification was the display to Glover of only a single picture, rather than a photo array. With good reason, such single-suspect procedures have "been widely condemned." Stovall v. Denno, 388 U.S., at 302, 87 S.Ct., at 1973. They give no assurance that the witness can identify the criminal from among a number of persons of similar appearance, surely the strongest evidence that there was no misidentification. In Simmons v. United States, our first decision involving photographic identification, we recognized the danger that a witness seeing a suggestively displayed picture will "retain in his memory the image of the photograph rather than of the person actually seen." 390 U.S., at 383-384, 88 S.Ct., at 971. "Subsequent identification of the accused then shows nothing except that the picture was a good likeness." Williams & Hammelmann, supra, n. 1, at 484. As Simmons warned, the danger of error is at its greatest when "the police display to the witness only the picture of a single individual . . . (and) is also heightened if the police indicate to the witness that they have other evidence that . . . the perso(n) pictured committed the crime." 390 U.S., at 383, 88 S.Ct., at 971. See also ALI, Model Code of Pre-Arraignment Procedure §§ 160.2(2), (5) (1975).

The use of a single picture (or the display of a single live suspect, for that matter) is a grave error, of course, because it dramatically suggests to the witness that the person shown must be the culprit. Why else would the police choose the person? And it is deeply ingrained in human nature to agree with the expressed opinions of others particularly others who should be more knowledgeable when making a difficult decision.15 In this case, moreover, the pressure was not limited to that inherent in the display of a single photograph. Glover, the identifying witness, was a state police officer on special assignment. He knew that D'Onofrio, an experienced Hartford narcotics detective, presumably familiar with local drug operations, believed respondent to be the seller. There was at work, then, both loyalty to another police officer and deference to a better-informed colleague.16 Finally, of course, there was Glover's knowledge that without an identification and arrest, government funds used to buy heroin had been wasted.

The Court discounts this overwhelming evidence of suggestiveness, however. It reasons that because D'Onofrio was not present when Glover viewed the photograph, there was "little pressure on the witness to acquiesce in the suggestion." Ante, at 116. That conclusion blinks psychological reality.17 There is no doubt in my mind that even in D'Onofrio's absence, a clear and powerful message was telegraphed to Glover as he looked at respondent's photograph. He was emphatically told that "this is the man," and he responded by identifying respondent then and at trial "whether or not he was in fact 'the man.' " Foster v. California, 394 U.S., at 443, 89 S.Ct., at 1129.18

I must conclude that this record presents compelling evidence that there was "a very substantial likelihood of misidentification" of respondent Brathwaite. The suggestive display of respondent's photograph to the witness Glover likely erased any independent memory that Glover had retained of the seller from his barely adequate opportunity to observe the criminal.

Since I agree with the distinguished panel of the Court of Appeals that the legal standard of Stovall should govern this case, but that even if it does not, the facts here reveal a substantial likelihood of misidentification in violation of respondent's right to due process of law, I would affirm the grant of habeas corpus relief. Accordingly, I dissent from the Court's reinstatement of respondent's conviction.