Even as DNA evidence revolutionizes criminal justice, courts and experts are beginning to question the value of a simpler form of identification made famous a century ago by Scotland Yard: the fingerprint.

A recent federal court ruling in a murder-for-hire trial in Philadelphia barred experts from telling the jury that two fingerprints definitively matched, concluding that the hallowed crime-fighting tool is more a skill subject to human error than science.

As prosecutors in the case warn of a potentially profound impact on the ability to convict criminals and try to overturn the decision, defense lawyers in Chicago and elsewhere say they see an opportunity to shoot holes in evidence once considered iron-clad.

"I think the case will have an impact," said Gregory O'Reilly, forensics supervisor of the Cook County public defender's office. "It will be an invitation to litigate these [fingerprint] issues and not to take a lab report at face value."

The ruling last month by U.S. District Judge Louis Pollak didn't exclude fingerprint evidence. Experts can still testify about similarities between two prints, but they can't tell jurors outright that a print belongs to a defendant.

Pollak concluded that such testimony isn't based firmly on science, because a single standard doesn't exist for what constitutes a "match" and a definitive statistical study hasn't been done to show how often examiners might be wrong. The ruling essentially puts fingerprint evidence in the same category as other less specific forensic evidence, such as shoe prints, said Terry Kiely, a law professor at DePaul University, who specializes in forensic evidence.

The ruling isn't binding in Illinois, but legal experts say it raises issues that judges will have to confront.

"This case will encourage public defenders all over the country to start filing motions," Kiely said.

In Cook County, a spokesman for State's Atty. Richard Devine's office said they expect to fend off the ruling, in part because it relied on a 1993 U.S. Supreme Court decision on scientific testimony that directly applies only to federal courts. But others contend that the landmark 1993 decision, seen as the court's way of trying to keep "junk science" out of medical and product liability lawsuits, has influenced how judges think about science, even in states like Illinois, which hasn't adopted it as precedent.

The opinion directed federal judges to look closely at whether scientific opinions have been tested, peer reviewed and studied, so the error rate, if any, is known.

Terence MacCarthy, the federal public defender in Chicago, calls Pollak's opinion a wake-up call, recalling courtroom battles in which FBI agents presented previously unassailable fingerprint evidence.

"When they finished it sounded like it was coming from Mt. Sinai," MacCarthy said. "We never made any inroads in it. ... This opinion changes all that."

Still, fingerprint skeptics can point to only a few cases where an incorrect match led to a false accusation or arrest. A Philadelphia man spent two years in prison for murder before being released in 1999, after the FBI found that the fingerprint match used to convict him had been an error.

In 1994 police investigating the still unsolved murder of seven people at Brown's Chicken & Pasta in Palatine questioned a man from Chicago's West Side for three days based on what a Chicago police examiner thought was a fingerprint match. The Illinois State Police and FBI later overturned the match, said David Grieve, latent print coordinator at the state police's Southern Illinois Forensic Science Centre in Carbondale.

British Import

Fingerprints have been a recognized tool of law enforcement for more than 90 years. The technique began to gain currency in the U.S. after a Scotland Yard detective presented a paper on the subject at the 1904 World's Fair in St. Louis.

Fingerprints are deposits of sweat left behind when the ridges of a person's fingertips touch a surface. To reveal prints at a crime scene or on a weapon, police may sprinkle a surface with powder or expose an object to fumes from heated glue.

The "latent" prints lifted from a crime scene prints can be compared with the "rolled" prints taken from a suspect. Or, if they have no suspect, police can run the prints through the FBI's Integrated Automated Fingerprint Identification System, a computerized database of about 43 million prints.

In the case before Pollak, federal prosecutors have charged three men with a series of drug-related killings in Philadelphia and Puerto Rico. Prosecutors want FBI and police experts to testify that prints found in a car match one of the defendants and a cooperating witness. The case is set for trial March 18.

No clear `match' standards

The fingerprint controversy stems from the fact that prints lifted from a crime scene are typically small--only about 22 percent the size of rolled prints, according to an FBI expert in the Philadelphia case--and are often smudged or distorted.

Trained fingerprint examiners search the latent prints in hope of finding enough similarities to declare a match with a rolled print. But how much similarity is enough to confidently declare a match? That judgment is left to each individual.

FBI fingerprint examiners try to find 12 common "Galton points," minute characteristics on the fingerprint ridges, before declaring a match. But the examiners can declare a match with fewer similarities, an FBI expert testified.

The lack of a uniform standard concerned Pollak, who cited the expert testimony of Dr. David Stoney, director of the McCrone Research Institute, a Chicago microscopy lab known for its work on historical artifacts such as the Shroud of Turin and hair samples from Napoleon Bonaparte and Ludwig van Beethoven.

Stoney said letting each examiner decide when a match exists means matching "is not a scientific determination."

Most expert testimony requires the witness to make some judgment call, argues Grieve, the state police expert.

In Illinois, matches are "based on each examiner's experience," said Jan Girten, section chief at the state police lab in Chicago and a print examiner for 10 years.

It's a judgment call the examiners are trained to make, Girten said.

In Illinois and elsewhere, each print is either found to be a match, not a match or unsuitable for a comparison. "There's no 99 percent," Girten said. "It's 100 percent one way or it's not."

That answer makes fingerprint skeptics bristle. No process is error free, they argue.

Joseph P. Polski, an official with the International Association for Identification, the main trade association for fingerprint examiners, concedes that no written standard exists.

"You can make an academic argument that this doesn't work," Polski said. "But we've proved over and over again that it does work."