Keith Allen Harward, center, hugs a well wisher, Rhonda Rowland, of Farmville, Va, as he is released from Nottoway Correctional Center in Burkeville, Va. (Daniel Sangjib Min/AP)

The exoneration of a former sailor who served 33 years for a 1982 rape and murder in Virginia has highlighted mounting concerns about the use of bite marks as evidence in criminal cases.

Keith Allen Harward, 60, was released Friday from Virginia’s Nottoway Correctional Center, where he had been serving a life sentence. Harward was convicted after six experts concurred that bite marks on one victim’s leg matched Harward to a “medical certainty.” But DNA testing last month proved those experts wrong — and cleared Harward.

“I had a life. I got locked up, now I got to start life all over,” Harward said in remarks made outside prison and streamed live by Richmond television station NBC12. He added: “They weren’t looking for the truth. They were looking for a conviction. . . . They need to stop this stuff.”

Harward was convicted in the beating death of Jesse Perron and the sexual assault on Perron’s wife in their home blocks from the Newport News naval shipyard on Sept. 14, 1982. The woman said the couple’s attacker was wearing Navy clothing, but she did not identify Harward.

On Thursday, Virginia’s Supreme Court declared Harward innocent and ordered his release. The move came after Attorney General Mark R. Herring (D) joined a defense motion citing DNA testing that identified the perpetrator as another sailor from the U.S.S. Carl Vinson. That man later died in prison in Ohio while being held for abduction, attempted burglary and other charges.

Keith Allen Harward, convicted of a 1982 murder and rape in Newport News, Virginia, is seen in an undated picture released by the Virginia Department of Corrections. (Handout/Reuters)

Forensic odontologists said Harward’s exoneration raises hard questions for a field whose validity has been undermined by scientific and legal developments. At least 25 people arrested or convicted on evidence that included bite marks have been exonerated by DNA, according to the Innocence Project.

Still, no state or federal appellate court has found bite-mark evidence to be inadmissible.

The Innocence Project, part of Harward’s legal team, called such evidence “grossly unreliable” and urged judges to halt its use in criminal prosecutions. “How many more Mr. Harwards do we have to find before courts take seriously the obligation to eliminate this unreliable evidence from being used when life and liberty are at stake?” said M. Chris Fabricant, director of strategic litigation for the project. “Every state in the nation should be conducting reviews to see if there are others like Mr. Harward sitting in prison for crimes they didn’t commit.”

In an interview, Adam Freeman, president of the American Board of Forensic Odontology, which accredits forensic dentists and sets guidelines, called the case “very troubling.” He said Harward was convicted using practices that were “state of the art” at the time.

Freeman said the field has a duty to correct any similar mistakes. “I would encourage all to review our individual cases,” he said, citing a recent message to the board’s 90 certified and roughly 30 actively testifying members, as well as an unknown number of noncertified experts, who consult on more than 100 cases a year, including capital cases. “We need to take this responsibility seriously.”

Testimony at Harward’s trial came from one of the field’s leading practitioners, Lowell J. Levine, then an independent consultant whose involvement in cases including serial killer Ted Bundy and Nazi war criminal Josef Mengele established his worldwide reputation. He has served as president of the American Academy of Forensic Sciences and the odontology board.

Keith Allen Harward, center right, waves he walks with Innocence Project's, from left, Dana Delger, Chris Fabricant and Olga Akselrod, as he is released from Nottoway Correctional Center in Burkeville, Va. (Daniel Sangjib Min/AP)

Levine testified to “a very, very, very high degree of probability those teeth left that bite mark,” referring to Harward, and said a coincidental match to another person was a “practical impossibility.”

Levine was joined by another government witness, Virginia dentist Alvin G. Kagey, who independently examined the evidence and testified “with all medical certainty” that “there is just not anyone else that would have this unique dentition.”

Two experts contacted by Harward’s trial defense concurred with Kagey and Levine and were not called to testify, according to limited defense records maintained by his family.

Another two dentists first ruled out Harward in the investigation while screening dental records of sailors aboard the ship, but they reversed themselves and corroborated the match to Newport News police, case records show.

In a statement, Levine said he was upset and disturbed at the result, given that the case produced wide agreement among examiners who had access to more information than is often available. “This case should persuade all my colleagues to agree with the need for more scientific research and investigation,” Levine said.

Harward’s release comes at a time of transition for bite-mark evidence in courts and greater scrutiny of forensic testimony.

In February, the Texas Forensic Science Commission became the first official state or federal entity to call for a moratorium in the use of bite-mark identification in criminal trials, saying the validity of the technique had not been scientifically established.

The odontology board also issued new guidelines that say experts should no longer make identifications but instead limit testimony to including or excluding a person as a potential contributor or declaring results inconclusive.

Many prosecutors and practitioners say skilled examiners exercising proper caution can glean valuable evidence in cases, including violent crimes where there is a limited pool of suspects and DNA is less helpful because the perpetrator is known to the victim, such as in child-abuse cases.

“Bite-mark comparisons can often serve as important corroborative evidence on very complex and serious cases involving vulnerable victims,” said Joan Vollero, spokeswoman for Manhattan District Attorney Cyrus R. Vance.

Vollero declined further comment, citing the pending prosecution of Clarence Dean in the death of an aspiring model in New York in 2007. The office withdrew bite-mark evidence in January to speed proceedings for witnesses, the victim’s family and the court in the case, after the Innocence Project renewed efforts to exclude that evidence.

The Texas commission cited research last year by Freeman’s group that cast doubt on fundamental assumptions of dental analysis. Freeman’s group found that board-certified members shown photos from real cases could not agree, in most instances, even whether injuries were caused by human teeth.