In 1994, when Monica Reyes was strangled, beaten, and left dead in an Elizabeth, NJ drainage ditch, the only evidence investigators found to identify her killer was a bite mark on her back. There were no witnesses. There was no CCTV footage. After examining the bite mark, a forensic odontologist identified Gerard Richardson as the killer.

Forensic odontologists are dentists who go through rigorous training after dental school to match bite marks to specific people. The process — established in the US by the American Board of Forensic Odontology (ABFO) — involves observing autopsies, coordinating with medical examiners, and participating in field exercises like "facial dissection[s] performed in conjunction with a multiple fatality incident." The goal is to understand human teeth and jaws as calling cards, identifiers that can be used in the same way as fingerprints in criminal cases. In 1995, based solely on an odontologist’s testimony, a jury found Gerard Richardson guilty of murder. He was sentenced to 30 years in prison.

When Richardson appealed the conviction, a protracted legal process ensued — one in which the Innocence Project of New York (IP) became intimately involved. IP, an organization affiliated with Yeshiva University’s law school that’s dedicated to probing wrongful convictions, has been pushing for a DNA test of the evidence in Richardson's case since 2006. This year, IP was finally successful. The DNA test revealed that Richardson’s genetic information was nowhere to be found. And last week, nearly two decades after Reyes’ death, IP filed a brief to throw Richardson’s conviction out of court.

Criminal appeals tend to move glacially and unpredictably, so it’s unclear if the court will exonerate Richardson or not. (Prosecutors have less than a month to prove that Richardson wasn’t incarcerated for nothing.) But IP says his case is not unique. The organization has already identified or helped to exonerate innocent people in 17 similar situations, and flagged seven cases in which flawed bite mark evidence was initially used to indict innocent people, but later squelched in court. Based on the number of cases tied to bite mark analysis — and IP’s belief that it has only intervened in a few of the ensuing wrongful convictions — the organization argues that the practice is "junk science" and should be abolished. Chris Fabricant, IP’s director of strategic litigation, says that bite mark evidence is "based on subjective speculation" and that the technique is "responsible for more wrongful convictions than any other forensic discipline that’s still in practice today."

Fabricant is just one of many people and organizations questioning the use of bite mark evidence in court.

Perhaps the most prominent of these organizations is the National Academy of Sciences (NAS). The prestigious, 2,200-member nonprofit voiced similar concerns in a 328-page report four years ago: "The simple reality is that the interpretation of forensic evidence [such as bite mark analysis] is not always based on scientific studies to determine its validity," the report reads. "This is a serious problem." Bolstering this point, the NAS report quoted a 2001 analysis from the peer-reviewed academic journal Science & Justice, concluding that there is a distinct "lack of valid evidence to support many of the assumptions made by forensic dentists during bite mark comparisons."

"At the very very least," Fabricant said, "we need to take a step back, stop admitting this [evidence in court], and begin doing some applied research to see if the claims that these dentists are making have any validity whatsoever."

But that’s a taller order than one might suspect.

Earlier this year, IP got involved in the case of Clarence Dean, who was convicted in a 2007 murder at the notoriously seedy Hotel Carter, about a block from Times Square. Dean — a drifter with multiple sex offenses on his record who allegedly stole a car to get to Manhattan and trolled online sex sites once he arrived — was convicted largely on the basis of bite mark analysis.

IP filed a brief earlier this year arguing that bite mark evidence should be disallowed in Dean’s case, and in all New York State cases going forward. This case was particularly important for IP, and for the larger movement against bite mark evidence: it was the first bite mark appeal to be argued in court since the 2009 NAS report, meaning that a US judge could consider the published opinion of a prestigious group of scientists before making a judgement. If a judge ruled against bite mark analysis in Dean’s case, it could set a precedent against such evidence in future cases nationwide.

Not all experts share IP's perspective on bite marks, though. Dr. Gregory Golden, president of AFBO, is one of them. He’s a longtime courtroom expert who has worked on hundreds of bite mark cases — including a few cases, pro bono, for the Innocence Project.

Speaking for himself — and not in his official capacity as ABFO’s president — Golden emphasized that there are hundreds of cases where bite mark evidence has led to reliable convictions (including serial killer Ted Bundy’s), and that there are circumstances where forensic odontology can help assess age, identify victims in mass casualties, and spot child abuse, which, Golden points out, often saves lives. Ideally, DNA evidence would be available at every crime scene, but Golden argues that bite marks are sometimes the sole forensic clue.

Golden concedes that there’s little scientific research to back claims from forensic odontologists in court — but he hopes to see that change. "What we’re trying to do," he says, "is to develop proper, unbiased research techniques that take into consideration real-time mechanisms or setups for researching bite marks."

The problem, he says, is that it’s difficult to conduct realistic studies on how bite marks injure living human flesh. In the past, studies have been conducted on cadavers and anesthetized pigs, with dental models mounted in vice grips. But such studies don’t accurately reflect bites on living human flesh, and Golden adds that "it’s almost impossible to find voluntary subjects offering themselves to be bitten severely enough to be wounded."

The question, then, is this: if bite mark evidence is so controversial, and burdened by such hurdles to scientific experimentation, why is it still allowed in court?

Largely, it’s because courts still consider bite mark analysis a valuable asset in some cases. "As gatekeepers, judges are aware of the precedents throughout the United States where bite mark evidence has been crucial to the jury reaching an accurate final verdict," Golden says. "Those of us who are experienced in bite mark analysis are intellectually honest, and have our egos in check and usually get it right."

No data exists to quantify how many US criminal cases have included bite mark evidence. But Golden estimates that such evidence has been correctly used to convict offenders "in nearly 1,000 cases," and says that forensic odontologists offer up accurate analysis about 98 percent of the time.

But the need for such approximations is exactly why advocates like Fabricant are pushing to disallow bite mark evidence until more research and data exist. "There’s a universe of people in prison who were put there by bite mark evidence," Fabricant says. "This is a flawed discipline, and it’s done immeasurable harm to untold numbers of wrongfully convicted people as well as the justice system itself."

Disclosure: From 2009 through 2011, Matt Stroud worked as a reporter and investigator for a university program in Pittsburgh loosely tied to the Innocence Project.