delivered the opinion of the Court, except as to Part II-D.
We are called upon to decide whether in the circumstances of this case an in-court identification of the accused by the victim of a crime should be suppressed as the fruit of the defendant’s unlawful arrest.
I
On the morning of January 3, 1974, a woman was accosted and robbed at gunpoint by a young man in the women’s restroom on the grounds of the Washington Monument. Her assailant, peering at her through a 4-inch crack between the wall and the door of the stall she occupied, asked for $10 and demanded that he be let into the stall. When the woman refused, the robber pointed a pistol over the top of the door and repeated his ultimatum. The victim then surrendered the money, but the youth demanded an additional $10. When the woman opened her purse and showed her assailant that she had no more cash, he gained entry to her stall and made sexual advances upon her. She tried to resist and pleaded with him to leave. He eventually did, warning his victim that he would shoot her if she did not wait at least 20 minutes before following him out of the restroom. The woman complied, and upon leaving the restroom 20 minutes later, immediately reported the incident to the police.
On January 6, two other women were assaulted and robbed in a similar episode in the same restroom. A young man threatened the women with a broken bottle, forced them to hand over $20, and then departed, again cautioning his victims not to leave for 20 minutes. The description of the *466robber given to the police by these women matched that given by the first victim: All three described their assailant as a young black male, 15-18 years old, approximately 5'5" to 5'8" tall, slender in build, with a very dark complexion and smooth skin.
Three days later, on January 9, Officer David Rayfield of the United States Park Police observed respondent in the area of the Washington Monument concession stand and restrooms. Aware of the robberies of the previous week and noting respondent’s resemblance to the police “lookout” that described the perpetrator, the officer and his partner approached respondent.1 Respondent gave the officers his name and said that he was 16 years old. When asked why he was not in school, respondent replied that he had just “walked away from school.” 2 The officers informed respondent of his likeness to the suspect’s description, but there was no further questioning about those events. Respondent was allowed to leave, and the officers watched as he entered the nearby restrooms.
While respondent was still inside, Officer Rayfield saw and spoke to James Dickens, a tour guide who had previously reported having seen a young man hanging around the area of the Monument on the day of the January 3d robbery. In response to the officer’s request to observe respondent as he left the restroom, Dickens tentatively identified him as the individual he had seen on the day of the robbery.
On the basis of this additional information, the officers again approached respondent and detained him. Detective Earl Ore, the investigator assigned to the robberies, was immediately summoned. Upon his arrival some 10 or 15 minutes later, Detective Ore attempted to take a Polaroid photo*467graph of respondent, but the inclement weather conditions frustrated his several efforts to produce a picture suitable for display to the robbery victims. Respondent was therefore taken into custody, ostensibly because he was a suspected truant. He was then transported to Park Police headquarters, where the police briefly questioned him, obtained the desired photograph, telephoned his school, and released him. Respondent was never formally arrested or charged with any offense, and his detention at the station lasted no more than an hour.
On the following day, January 10, the police showed the victim of the first robbery an array of eight photographs, including one of respondent. Although she had previously viewed over 100 pictures of possible suspects without identifying any of them as her assailant, she immediately selected respondent’s photograph as that of the man who had robbed her. On January 13, one of the other victims made a similar identification.3 Respondent was again taken into custody, and at a court-ordered lineup held on January 21, he was positively identified by the two women who had made the photographic identifications.
The grand jury returned an indictment against respondent on February 22, 1974, charging him with two counts of armed robbery, two counts of robbery, one count of attempted armed robbery, and three counts of assault with a dangerous weapon.4 Respondent filed a pretrial motion to suppress all identification testimony, contending that his detention on the truancy charges had been merely a pretext to allow the police to obtain evidence for the robbery investigation. After hearing extensive testimony from the three victims, the police officers, and respondent, the trial court found that the respondent’s detention at Park Police headquarters on January 9 consti*468tuted an arrest without probable cause.5 Accordingly, the court ruled that the products of that arrest — the photographic and lineup identifications — could not be introduced at trial. But the judge concluded that the victims’ ability to identify respondent in court was based upon independent recollection untainted by the intervening identifications, and therefore held such testimony admissible. At trial, all three victims identified respondent as their assailant. On April 23, the jury convicted him of armed robbery of the first victim, but returned verdicts of not guilty on all other charges.6 Respondent was sentenced to four years’ probation under the Federal Youth Corrections Act, 18 U. S. C. § 5010 (a).
On appeal, the District of Columbia Court of Appeals, sitting en banc, reversed respondent’s conviction and ordered the suppression of the first robbery victim’s in-court identi*469fication.7 389 A. 2d 277 (1978). The court viewed its decision to be a wholly conventional application of the familiar “fruit of the poisonous tree” doctrine. See Wong Sun v. United States, 371 U. S. 471 (1963); Silverthorne Lumber Co. v. United States, 251 U. S. 385 (1920). After upholding the trial court’s finding that respondent was detained without probable cause — a determination that is not challenged in this Court8 — the Court of Appeals turned to consideration of what evidentiary consequences ought to flow from that Fourth Amendment violation. In deciding whether the in-court identification should have been suppressed, the court observed that the analysis must focus on whether the evidence was obtained by official “exploitation” of the “primary illegality” within the meaning of Wong Sun, supra,9 and that the principal issue was whether the unlawful police behavior bore a causal relationship to the acquisition of the challenged testimony. The court answered that question in the affirmative, reasoning that but for respondent’s unlawful arrest, the police would not have obtained the photograph that led to his subsequent identification by the complaining witnesses and, ultimately, prosecution of the case.10 Satisfied that the *470in-court identification was thus at least indirectly the product of official misconduct, the court then considered whether any of three commonly advanced exceptions to the exclusionary rule — the “independent source,” “inevitable discovery,” or “attentuation” doctrines11 — nonetheless justified its admission. Finding these exceptions inapplicable, the Court of Appeals concluded that, the in-court identification testimony should have been excluded as a product of the violation of respondent’s Fourth Amendment rights. We granted certiorari. 440 U. S. 907 (1979). We reverse.
II
Wong Sun, supra, articulated the guiding principle for determining whether evidence derivatively obtained from a violation of the Fourth Amendment is admissible against the accused at trial: “The exclusionary prohibition extends as well to the indirect as the direct products of such invasions.” 371 U. S., at 484. See Silverthome Lumber Co. v. United States, supra; Weeks v. United States, 232 U. S. 383 (1914). As subsequent cases have confirmed, the exclusionary sanction applies to any “fruits” of a constitutional violation — whether such evidence be tangible, physical material actually seized in an illegal search,12 items observed or words overheard in the course of the unlawful activity,13 or confessions or statements of the accused obtained during an illegal arrest and detention.14
*471In the typical “fruit of the poisonous tree” case, however, the challenged evidence was acquired by the police after some initial Fourth Amendment violation, and the question before the court is whether the chain of causation proceeding from the unlawful conduct has become so attenuated or has been interrupted by some intervening circumstance so as to remove the “taint” imposed upon that evidence by the original illegality. Thus most cases begin with the premise that the challenged evidence is in some sense the product of illegal governmental activity. It is the Court of Appeals’ application of that premise to the facts of this case that we find erroneous.
A ,victim’s in-court identification of the accused has three distinct elements. First, the victim is present at trial to testify as to what transpired between her and the offender, and to identify the defendant as the culprit. Second, the victim possesses knowledge of and the ability to reconstruct the prior criminal occurrence and to identify the defendant from her observations of him at the time of the crime. And third, the defendant is also physically present in the courtroom, so that the victim can observe him and compare his appearance to that of the offender. In the present case, it is our conclusion that none of these three elements “has been come at by exploitation” of the violation of the defendant’s Fourth Amendment rights. Wong Sun, supra, at 488.
A
In this case, the robbery victim’s presence in the courtroom at respondent’s trial was surely not the product of any police misconduct. She had notified the authorities immediately after the attack and had given them a full description of her assailant. The very next day, she went to the police station to view photographs of possible suspects, and she voluntarily assisted the police in their investigation at all times. Thus this is not a case in which the witness’ identity was discovered or her cooperation secured only as a result of an unlawful *472search or arrest of the accused.15 Here the victim’s identity was known long before there was any official misconduct, and her presence in court is thus not traceable to any Fourth Amendment violation.
B
Nor did the illegal arrest infect the victim’s ability to give accurate identification testimony. Based upon her observations at the time of the robbery, the victim constructed a mental image of her assailant. At trial, she retrieved this mnemonic representation, compared it to the figure of the defendant, and positively identified him as the robber.16 No part of this process was affected by respondent’s illegal arrest. In the language of the “time-worn metaphor” of the poisonous tree, Harrison v. United States, 392 U. S. 219, 222 (1968), the toxin in this case was injected only after the evidentiary bud had blossomed; the fruit served at trial was not poisoned.
This is not to say that the intervening photographic and lineup identifications — both of which are conceded to be suppressible fruits of the Fourth Amendment violation — could not under some circumstances affect the reliability of the in-court identification and render it inadmissible as well. Indeed, given the vagaries of human memory and the inherent suggestibility of many identification procedures,17 just *473the opposite may be true. But in the present case the trial court expressly found that the witness’ courtroom identification rested on an independent recollection of her initial encounter with the assailant, uninfluenced by the pretrial identifications, and this determination finds ample support in the record.18 In short, the victim’s capacity to identify her assailant in court neither resulted from nor was biased by the unlawful police conduct committed long after she had developed that capacity.19
*474c
Insofar as respondent challenges his own presence at trial, he cannot claim immunity from prosecution simply because his appearance in court was precipitated by an unlawful arrest. An illegal arrest, without more, has never been viewed as a bar to subsequent prosecution, nor as a defense to a valid conviction. Gerstein v. Pugh, 420 U. S. 103, 119 (1975); Frisbie v. Collins, 342 U. S. 519 (1952); Ker v. Illinois, 119 U. S. 436 (1886).20 The exclusionary principle of Wong Sun and Silverthorne Lumber Co. delimits what proof the Government may offer against the accused at trial, closing the courtroom door to evidence secured by official lawlessness. Respondent is not himself a suppressible “fruit,” and the illegality of his detention cannot deprive the Government of the opportunity to prove his guilt through the introduction of evidence wholly untainted by the police misconduct.
D*
Respondent argues, however, that in one respect his corpus is itself a species of “evidence.” When the victim singles out respondent and declares, “That’s the man who robbed me,” his physiognomy becomes something of evidentary value, much like a photograph showing respondent at the scene of the *475crime.21 And, as with, the introduction of such a photograph, he contends that the crucial inquiry for Fourth Amendment purposes is whether that evidence has become available only as a result of official misconduct. We read the Court of Appeals’ opinion as essentially adopting this analysis to support its suppression order. See 389 A. 2d, at 285-287.
We need not decide whether respondent’s person should be considered evidence, and therefore a possible “fruit” of police misconduct. For in this case the record plainly discloses that prior to his illegal arrest, the police both knew respondent’s identity and had some basis to suspect his involvement in the very crimes with which he was charged. Moreover, before they approached respondent, the police had already obtained access to the “evidence” that implicated him in the robberies, i. e., the mnemonic representations of the criminal retained by the victims and related to the police in the form of their agreement upon his description. In short, the Fourth Amendment violation in this case yielded nothing of evidentiary value that the police did not already have in their grasp.22 Rather, respondent’s unlawful arrest served merely to link together two extant ingredients in his identification. The exclusionary rule enjoins the Government from benefiting from evidence it has unlawfully obtained; it does not reach backward to taint information that was in official hands prior to any illegality.
Accordingly, this case is very different from one like Davis v. Mississippi, 394 U. S. 721 (1969), in which the defendant’s identity and connection to the illicit activity were only first discovered through an illegal arrest or search. In that case, the defendant’s fingerprints were ordered suppressed as the *476fruits of an unlawful detention. A woman had been raped in her home, and during the next 10 days, the local police rounded up scores of black youths, randomly stopping, interrogating, and fingerprinting them. Davis’ prints were discovered to match a set found at the scene of the crime, and on that basis he was arrested and convicted. Had it not been for Davis’ illegal detention, however, his prints would not have been obtained and he would never have become a suspect.' Here, in contrast, the robbery investigation had already focused on respondent, and the police had independent reasonable grounds to suspect his culpability.
We find Bynum v. United States, 104 U. S. App. D. C. 368, 262 F. 2d 465 (1958), cited with approval in Davis, supra, at 724, helpful in our analysis as well. In Bynum, the defendant voluntarily came down to the police station to look for his brother, who had been arrested earlier that day while driving an auto sought in connection with a robbery. After telling one of the officers that he owned the car, Bynum was arrested and fingerprinted. Those prints were later found to match a set at the scene of the robbery, and Bynum was convicted based in part on that evidence. The Court of Appeals held that the police lacked probable cause at the time of Bynum’s arrest, and it ordered the prints suppressed as “something of evidentiary value which the public authorities have caused an arrested person to yield to them during illegal detention.” 104 U. S. App. D. C., at 370, 262 F. 2d, at 467. As this Court noted in Davis, however, 394 U. S., at 725-726, n. 4, Bynum was subsequently reindicted for the same offense, and the Government on retrial introduced an older set of his fingerprints, taken from an FBI file, that were in no' way connected with his unlawful arrest. The Court of Appeals affirmed that conviction, holding that the fingerprint identification made on the basis of information already in the FBI’s possession was not tainted by the subsequent illegality and was therefore admissible. Bynum v. United States, 107 U. S. App. D. C. 109, 274 F. 2d 767 (1960).
*477The parallels between Bynum and this case are apparent: The pretrial identification obtained through use of the photograph taken during respondent’s illegal detention cannot be introduced; but the in-court identification is admissible, even if respondent’s argument be accepted, because the police’s knowledge of respondent’s identity and the victim’s independent recollections of him both antedated the unlawful arrest and were thus untainted by the constitutional violation. The judgment of the Court of Appeals is accordingly
Reversed.
Mr. Justice Marshall took no part in the consideration or decision of this case.Officer Rayfield testified that his suspicions were further aroused both by respondent’s presence on the almost deserted park grounds and by his apparently aimless meanderings around the restroom and concessions area.
Tr. 52. References are to the transcript of the suppression hearing and trial held on April 22 and 23, 1974, in the Superior Court of the District of Columbia.
The third victim did not review the photographic array, nor did she attend the subsequent lineup.
See D. C. Code §§ 22-502, 22-2901, and 22-3202 (1973).
The suppression hearing produced conflicting testimony as to the reasons for the attempt to photograph respondent. Officer Rayfield asserted that respondent was processed as a routine juvenile truant, a procedure that involves photographing the suspect and then calling his school and home to determine whether he is in fact truant. Tr. 53-54. Rayfield did acknowledge, however, that he had some suspicion that respondent was the robber described in the police description. Id., at 55, 57. Similarly, Detective Ore, while maintaining that respondent was apprehended and taken down to Park Police headquarters as a suspected truant, id., at 61, 63, admitted that his intent in trying to photograph him was to obtain a picture that could be shown to the complaining witnesses. Id., at 59.
The Government does not now attempt to justify respondent’s detention on the truancy charge, nor did it raise that argument in the court below. The Court of Appeals found that the procedures followed in respondent’s case did not conform to the typical truancy practices described by the police and that the officers never even superficially pursued the truancy matter. By the same token, the court expressly disavowed the existence of a “sham” or “pretext” arrest, and it analyzed respondent’s apprehension as a traditional arrest for armed robbery and assault without probable cause. 389 A. 2d 277, 299-300, n. 32 (DC 1978).
Because respondent was acquitted of all charges in connection with the robberies of January 6, the only issue raised on his appeal was the admissibility of the first robbery victim’s in-court identification.
On February 16, 1977, a division of the Court of Appeals originally affirmed respondent’s conviction, 369 A. 2d 1063. Three months later, however, the full court granted respondent’s motion for rehearing and vacated its earlier judgment. Record 356.
See Brief for United States 5, n. 4.
“We need not hold that all evidence is ‘fruit of the poisonous tree’ simply because it would not have come to light but for the illegal actions of the police. Rather, the more apt question in such a ease is 'whether, granting establishment of the primary illegality, the evidence to which instant objection is made has been come at by exploitation of that illegality or instead by means sufficiently distinguishable to be purged of the primary taint.’ Maguire, Evidence of Guilt, 221 (1959).” Wong Sun v. United States, 371 U. S., at 487-488.
“[T]he unlawful arrest produced photographs which were shown to the complaining witnesses who, as a result, identified [respondent); this resulted in his reapprehension, which yielded a court-ordered lineup iden*470tification and, eventually, in-court identification testimony during prosecution of the case.” 389 A. 2d, at 289.
See Nardone v. United States, 308 U. S. 338, 341 (1939) (attenuation); Silverthome Lumber Co. v. United States, 251 U. S. 385, 392 (1920) (independent source); United States ex rel. Owens v. Twomey, 508 F. 2d 858, 865 (CA7 1974) (inevitable discovery).
E. g., Whiteley v. Warden, 401 U. S. 560 (1971); Sibron v. New York, 392 U. S. 40 (1968); Beck v. Ohio, 379 U. S. 89 (1964).
E. g., United States v. Giordano, 416 U. S. 505 (1974).; see Silverman v. United States, 365 U. S. 505 (1961); McGinnis v. United States, 227 F. 2d 598 (CA1 1955).
E. g., Dunaway v. New York, 442 U. S. 200 (1979); Brown v. Illinois, 422 U. S. 590 (1975).
See generally Ruffin, Out on a Limb of the Poisonous Tree: The Tainted Witness, 15 UCLA L. Rev. 32 (1967).
At oral argument, the Government compared the witness’ mental image to an undeveloped photograph of the robber that is given to the police immediately after the crime, but which becomes visible only at the trial. Tr. of Oral Arg. 11-12. Although this analogy may not comport precisely with current psychological theories of perception, see, e. g., Buckout, Eyewitness Testimony, Scientific American 23 (Dec. 1974), it is apt for purposes of analysis.
See, e. g., P. Wall, Eye-Witness Identification in Criminal Cases 40-64 (1965); Note,. Did Your Eyes Deceive You? Expert Psychological Testimony on the Unreliability of Eyewitness Identification, 29 Stan. L. Rev. 969, 974-989 (1977).
United States v. Wade, 388 U. S. 218 (1967), enumerated several factors for consideration in applying the “independent origins” test. Id., at 241. Cf. Manson v. Brathwaite, 432 U. S. 98 (1977); Neil v. Biggers, 409 U. S. 188 (1972). We attach particular significance to the following circumstances which support the trial court’s determination in this case: the victim viewed her assailant at close range for a period of 5-10 minutes under excellent lighting conditions and with no distractions, Tr. 4, 7, 111; respondent closely matched the description given by the victim immediately after the robbery, id., at 52, 59; the victim failed to identify anyone other than respondent, id., at 8, but twice selected respondent without hesitation in nonsuggestive pretrial identification procedures, id., at 9-11; and only a week had passed between the victim’s initial observation of respondent and her first identification of him, id., at 8-9.
Our reliance on the fact that the witness twice identified respondent in out-of-court confrontations is not intended to assign any independent evidentiary value to those identifications for to do so would undermine the exclusionary rule’s objectives in denying the Government the benefit of any evidence wrongfully obtained. Rather, the accurate pretrial identifications assume significance only to the extent that they indicate that the witness’ ability to identify respondent antedated any police misconduct, and hence that her in-court identification had an “independent source.”
Respondent contends that the “independent source” test of United States v. Wade, supra, and Stovall v. Denno, 388 U. S. 293 (1967), although derived from an identical formulation in Wong Sun, see 388 U. S., at 241, seeks only to determine whether the in-court identification is sufficiently reliable to satisfy due process, and is thus inapplicable in the context of this Fourth Amendment violation. We agree that a satisfactory resolution of the reliability issue does not provide a complete answer to the considerations underlying Wong Sun, but note only that in the present case both concerns are met.
Cf. United States v. Blue, 384 U. S. 251, 255 (1966):
“Our numerous precedents ordering the exclusion of such illegally obtained evidence assume implicitly that the remedy does not extend to barring the prosecution altogether. So drastic a step might advance marginally some of the ends served by exclusionary rules, but it would also increase to an intolerable degree interference with the public interest in having the guilty brought to book.”
In some cases, of course, prosecution may effectively be foreclosed by the absence of the challenged evidence. But this contemplated consequence is the product of the exclusion of specific evidence tainted by the Fourth Amendment violation and is not the result of a complete bar to prosecution.
This part is joined only by Mb. Justice Stewart and Mr. Justice Stevens.
Cf. Stevenson v. Mathews, 529 F. 2d 61, 63 (CA7 1976).
Thus we are not called upon in this ease to hypothesize about whether routine investigatory procedures would eventually have led the police to discover respondent’s culpability. His involvement in the robberies was already suspected, and no new evidence was acquired through the violation of his Fourth Amendment rights.