"The innocence list," a listing of 102 former inmates taken off of death rows across the country, may be the most powerful weapon fielded by opponents of capital punishment.

In 1998, a gathering of 28 of the once-condemned men and women in Chicago drew worldwide attention and cast doubt on the accuracy of state justice systems that impose death sentences.

Usually called the "exonerated," but sometimes the "innocent," the list was cited repeatedly by death-penalty opponents after then-Illinois Gov. George Ryan's decision last weekend to commute the death sentences of all 167 condemned inmates in the state.

But are all 102 truly innocent?

Many of those on the list, while removed from death row, clearly committed the crimes for which they faced execution, say some officials and capital-punishment supporters who have examined some or all of the cases.

The list, "Innocence: Freed From Death Row," is maintained by the Death Penalty Information Center in Washington. The organization opposes capital punishment as it is now practiced, said Richard Dieter, the director.

Dieter strongly defends the accuracy and validity of the list. While acknowledging that some on the list could have committed the crimes, they are all legally innocent, he said.

The most persistent critic of the list is Dudley Sharp of the Houston-based Justice for All, a victim-advocacy organization.

"No one disputes that innocents are found guilty, within all countries," Sharp conceded. But he said of the list of 102: "That number is a fraud."

Sharp said opponents have combined the factually innocent, such as those cleared by DNA, with those released because of legal errors, "thereby fraudulently raising the 'innocent' numbers."

Others on the list were prosecuted before the 1972 U.S. Supreme Court decision that forced a four-year halt on the death penalty, Sharp said. Those cases, he said, have no relevance to a discussion of innocence within the modern death-penalty framework.

To buttress his contentions, Sharp cites, among other studies, one performed last year by the Florida Commission on Capital Cases.

It looked at the 24 cases from Florida on the center's list and concluded that just "four of those 24 might be innocent - an 84 percent error rate in death-penalty opponents' claims."

Sharp argues that "if that error rate is consistent nationally, that would indicate that 17 of the alleged 102 innocents might be actually innocent." That would translate to a 0.2 percent error rate for the 7,300 people sentenced to death in the United States since 1973. "None were executed," he noted.

What the list proves, said Sharp, is that the system is good at weeding out the innocent.

John J. Kinsella, a prosecutor in the DuPage County, Ill., state's attorney's office, wrote in a report about Illinois cases on the list that "the true difficulty in the term 'exonerate' is that, despite its technical correctness ['to clear from accusation or blame'], its use may cause individuals to reach the conclusion that the underlying prosecution was 'wrongful' because the accused was in fact innocent.

"Such specious reasoning suggests that an acquittal or dismissal, under any circumstances, means the prosecution was somehow corrupt and flawed from its inception and the evidence did not warrant prosecution. Following the logic of this argument, anyone ever acquitted for any reason was in fact innocent and wrongfully prosecuted. This, of course, is not true," Kinsella wrote.

Here are some of the examples from the list and why Sharp takes issue with them: