The Central Park Five: What the 2002 Review Changed

The decisive change in the Central Park Five case came from the combination of a confession, forensic testing and a prosecutor willing to reconsider how the original evidence fit together. In 2002, Matias Reyes said that he alone had attacked and raped the woman jogging in Central Park in 1989. DNA testing connected him to semen recovered in the case, while no physical or forensic evidence connected Antron McCray, Kevin Richardson, Yusef Salaam, Raymond Santana or Korey Wise to that attack.

That discovery did not automatically erase the convictions. The five men filed post-conviction motions, the Manhattan District Attorney’s Office investigated the new evidence, and Justice Charles J. Tejada evaluated whether it probably would have produced more favorable verdicts. On December 19, 2002, he vacated all remaining convictions and ordered new trials. Prosecutors then dismissed the indictments rather than retrying the men.

The review therefore changed more than the status of the rape-related counts. It dismantled the legal judgments produced by the two 1990 trials and exposed how heavily the prosecution had depended on statements that lacked meaningful independent support.

What the original prosecution claimed

McCray, Richardson, Salaam, Santana and Wise were prosecuted under Indictment No. 4762/89. McCray, Salaam and Santana were tried together in 1990; Richardson and Wise faced a second jury trial. The indictment covered the attack on the woman jogging in Central Park as well as attacks on two male joggers and riot-related allegations. Some convictions had already been set aside at the original sentencing because of the defendants’ ages, but the remaining convictions continued to define the case until 2002.

According to the District Attorney’s later assessment, the trial case concerning the woman rested almost entirely on statements made by the five defendants. Those statements were powerful to jurors because they appeared to place the defendants within a group attack. But their evidentiary limits were substantial even before the new confession emerged.

None of the five admitted raping the jogger. Their accounts differed materially about the sequence of events, the people involved, weapons, clothing and the location of important actions. Such discrepancies do not necessarily make every part of a statement false. In this case, however, they mattered because the prosecution lacked physical evidence capable of independently placing the five at the attack on the woman.

Semen DNA recovered from a sock and a cervical swab did not match any of the five. The trial-era position was that the biological material could have come from an additional, unidentified participant. That possibility allowed the prosecution to argue that the exclusions did not clear the defendants. But exclusionary DNA cannot identify who did participate, and the unidentified-person theory depended on other evidence establishing that the five had been involved.

The later court record was stark: no physical or forensic evidence recovered from the victim, her belongings or the scene connected the five defendants to the attack or established how many perpetrators there had been. The statements were not one strand in a tightly corroborated case. They were its center of gravity.

Reyes’s confession changed the meaning of the DNA

In early 2002, Reyes, who was imprisoned for other offenses, said that he alone had attacked and raped the jogger. The Manhattan District Attorney’s Office learned of his account in February and began investigating. On May 8, the office was informed that an FBI comparison matched Reyes’s DNA to DNA recovered from the sock.

Further testing commissioned by prosecutors associated Reyes with the semen evidence and a pubic hair found on the sock. The testing identified him as the sole contributor to the DNA samples examined. This did not merely add another possible participant to the case. It identified the source of biological evidence that had previously excluded the five and supplied a concrete way to test Reyes’s account.

Prosecutors did not base their eventual position on the DNA match alone. They assessed Reyes’s description against objective details, examined whether he had a known connection to the five or their group in 1989, and considered features of other attacks he had committed. The office concluded that his account was corroborated by, or consistent with, objective evidence and found no known prior connection that would explain why he would falsely take responsibility on behalf of the five.

This distinction is essential. DNA established Reyes as the source of tested biological evidence; it did not, by itself, reconstruct every event in the park. His confession supplied an account. Investigators then examined whether the account and the physical record supported each other. The resulting combination forced a new evaluation of evidence that had once been interpreted through a group-attack theory.

The hair evidence could no longer carry its earlier implication

The 2002 review also revisited three hairs introduced against Richardson at trial. On re-evaluation, the hairs were found unsuitable for comparison and could not connect him to the jogger.

That correction illustrates why forensic language matters. A microscopic hair comparison is not a personal identification. The National Institute of Standards and Technology’s guidance on forensic trace comparisons warns that the number of people who could be included as possible sources through such a comparison is unknown. An apparent similarity therefore cannot be treated like a DNA identification, and an unsuitable hair cannot support even that limited comparison.

The problem was not simply that later science looked more sophisticated. The relevant question was what the earlier evidence had actually established. By 2002, the hair evidence could not provide the independent connection that the original statement-based case needed.

How the post-conviction motions reached the court

McCray, Richardson and Santana filed motions under New York Criminal Procedure Law Section 440.10(1)(g) on September 5, 2002. Salaam filed on September 9, and Wise followed on September 24. The consolidated motions relied on newly discovered evidence: Reyes’s sworn account and the forensic results connecting him to the biological evidence.

The governing provision of New York’s post-conviction statute allows a court to vacate a judgment when qualifying evidence discovered after trial could not have been produced earlier with due diligence and creates a probability that the verdict would have been more favorable to the defendant. It is not enough for new information to be interesting or merely cumulative. The court must assess its likely effect on the original verdict.

On December 5, 2002, the Manhattan District Attorney’s Office filed a 58-page affirmation joining the request to vacate. That position was consequential because prosecutors did not attempt to preserve the convictions by treating Reyes as an additional attacker. Instead, they concluded that the new evidence substantially weakened the original proof—especially the defendants’ statements, given their inconsistencies and the absence of meaningful independent evidence tying the five to the jogger attack.

Why the review reached charges beyond the jogger attack

A narrower review might have challenged only the convictions directly associated with the attack on the woman. The District Attorney’s Office took a broader position because the original prosecution had presented the night as a connected group incident and had obtained the defendants’ admissions about the different attacks through the same questioning process.

No victim in the other charged attacks identified any of the five. Once the new evidence undermined the prosecution’s central account of the jogger attack, prosecutors concluded that jurors would probably assess the remaining allegations differently as well. The statements could not be cleanly separated into a discredited section and a dependable section when they had been elicited together and used to support one overarching theory.

What Justice Tejada ordered

In the published decision in People v. Wise, Justice Tejada held that Reyes’s confession and the corroborating forensic evidence created a reasonable probability that the original verdicts would have been more favorable if jurors had heard the new evidence. On December 19, 2002, he granted the motions as to every remaining conviction.

The order vacated the convictions and directed new trials. It was not a sentence reduction, a pardon or an appellate ruling confined to the rape-related charges. Vacatur returned the cases to a posture in which prosecutors would have had to decide whether to try the charges again. They dismissed the indictments instead. The resulting criminal-case outcome was therefore vacatur of the convictions followed by dismissal of the charges.

A later NYPD-commissioned panel disagreed with the District Attorney’s conclusion that Reyes acted alone and expressed the view that the five could have participated. That assessment did not reverse Justice Tejada’s order, restore any conviction or produce a new prosecution.

The distinction between disagreement and legal outcome matters. A later panel could dispute the prosecutor’s reconstruction, but it could not make the vacated judgments valid again. The operative result remained that the convictions were set aside and the indictments dismissed.

Steven Lopez’s case was separate

Steven Lopez is sometimes folded into accounts of the Central Park Five, but he was not one of the five defendants whose convictions were reviewed in the 2002 motions. Lopez had been separately indicted and pleaded guilty in 1991 to robbing a male jogger. He did not go to trial, and his conviction was not included in Justice Tejada’s order.

His case changed much later. On July 25, 2022, the Manhattan District Attorney’s Office and Lopez jointly requested vacatur of his robbery conviction and dismissal of his indictment after a Post-Conviction Justice Unit review. The office cited the lack of reliable physical evidence, recanted statements implicating him, unreliable hair-comparison evidence and its conclusion that the circumstances made his plea involuntary. Chief Administrative Judge Ellen Biben granted the motion.

Keeping the cases separate preserves the actual chronology: five defendants obtained vacatur and dismissal in 2002 through consolidated newly discovered evidence motions; Lopez obtained relief in 2022 through a separate review of his plea and evidence.

The civil settlement did not create the 2002 result

The five men later pursued federal civil-rights litigation against New York City and other defendants. New York City approved a $41 million settlement in 2014, according to the City Comptroller’s office. That payment belonged to a separate civil proceeding. It did not vacate the convictions, dismiss the criminal charges or alter the reasoning of the 2002 judicial decision.

This sequence is often compressed into a single story of exoneration and compensation, but each step had a different legal function. The confession and forensic results supplied newly discovered evidence. The District Attorney’s review reassessed the original prosecution. Justice Tejada vacated the judgments. Prosecutors dismissed the indictments. The civil litigation came afterward and addressed different claims.

What the 2002 review ultimately established

The 2002 review changed the Central Park Five case because it made the original theory impossible to evaluate as though nothing new had happened. Reyes’s confession was supported by biological evidence. The same evidence excluded the five, while the statements against them materially conflicted and lacked meaningful physical corroboration. Hair evidence once presented against Richardson could not connect him to the victim. Prosecutors concluded that these changes weakened not only the jogger-attack counts but the connected prosecution as a whole.

The court’s response was legally precise and comprehensive: every remaining conviction was vacated, new trials were authorized, and the indictments were then dismissed. The men were not merely given shorter sentences, nor did the state preserve a subset of the original convictions.

The lasting lesson is about corroboration. A statement can direct an investigation, but its reliability must be tested against evidence that exists independently of it. In this case, the later forensic record did not reinforce the original accounts. It identified another man, supported his confession and exposed how little independent evidence had connected the five defendants to the crime at the center of their prosecution.