How Mistaken and Perjured Eyewitness Identification Testimony Put 46 Innocent Americans on Death RowErroneous eyewitness testimony – whether offered in good faith or perjured – no doubt is the single greatest cause of wrongful convictions in the U.S. criminal justice system.

Yet the extent of the problem – and, for that matter, the extent of any of the myriad other problems known to lead to wrongful convictions – can never be known, for the simple reason that it is visible only when innocence is clearly established; erroneous eyewitness identifications in cases in which innocence cannot be proved are permanently out of sight.

Nor is it possible to accurately gauge the problem even in the sizable body of cases that prosecutors abandon or in which there are acquittals after reversals on appeal. One problem is that there is no systematic way to identify all of the cases. While appellate court decisions are published and amenable to computer searches, dismissed charges and trial court acquittals are no where systematically catalogued, or at least not in data bases accessible to the public. Another problem is that cases often end without definitive resolutions. Were the eyewitnesses wrong – or lying – or was the case dropped or lost merely because the evidence was insufficient to prove guilt beyond a reasonable doubt? Who is to say whether the defendant is actually innocent or just got off on a technicality?

There are, however, two groups of cases in which the problem can be assessed with a reasonable – although certainly not absolute – degree of precision. The first group is cases in which innocence has been established through DNA. In June of 2000, the Center on Wrongful Convictions analyzed the first group and found that, of 67 then-known DNA exonerations in the United States and Canada, 51 of the convictions – 76.1% – had been based in whole or part on eyewitness identification testimony. The second group – the object of this analysis¹ – is capital cases in which the defendants have been exonerated.

Of course, we make no claim of 100% accuracy – an assessment of actual innocence is necessarily subjective. What we have attempted to do is provide an accurate, intellectually honest description of every relevant case – and we invite anyone aware of errors or additional facts that might put a case in a different light to call such information to our attention.


The Center on Wrongful Convictions identified and analyzed the cases of 86 defendants(84men and two women)who had been sentenced to death but legally exonerated based on strong claims of actual innocence since capital punishment was restored following the U.S. Supreme Court’s 1972 decision in Furman v. Georgia.

The analysis shows:

– Of the 86 legally exonerated persons, eyewitness testimony played a role in the convictions of 46 – 53.5%.

– Eyewitness testimony was the only evidence against 33 defendants – 38.4%.

– Only one eyewitness testified in 32 of the 46 defendants’ cases – 69.6% – and multiple eyewitnesses testified in the other 14 – 30.4%.

– The eyewitnesses were strangers to 19 of the defendants – 41.3% – and were non-accomplice acquaintances of 9 – 19.6%.

– Witnesses who presented themselves as combination accomplice-eyewitnesses testified against 15 of the defendants – 32.6% of the 46 – and all of these witnesses had incentives to testify, ranging from full immunity to leniency in sentencing.

– In five cases – each involving a single, non-accomplice eyewitness – the witness received consideration from the prosecution in a pending case.

– In four cases – each also involving a single, non-accomplice eyewitness – the apparently false testimony appeared to have been motivated by a grudge; in one case, the defendant and the purported eyewitness were in a love triangle.

– Eyewitness testimony was by far the most ubiquitous factor in the cases of the 86 defendants, followed by police and prosecutorial misconduct in the cases of 17 defendants (19.8%), jailhouse informant testimony in the cases of 10 (11.6%), so-called junk science in 9 (10.5%), false or coerced confessions in eight (9.3%), and various miscellaneous factors, such as questionable circumstantial evidence and hearsay, in 29 (33.7%).

– Of the 13 defendants’ cases in which the eyewitness identification was not the sole factor leading to the conviction, police or prosecutorial misconduct was the most significant other factor in seven defendants’ cases, jailhouse informant testimony in two, and false confession in one. Multiple factors were involved in the remaining three defendants’ cases – a combination of official misconduct and informant testimony in one, of junk science and hearsay testimony in another, and of misconduct and misleading circumstantial evidence in the third.

– The average (mean) time between the arrest of the defendant and his or her exoneration in the eyewitness cases was 95 months – just short of 8 years.