In 2004, a New Jersey prosecutor announced that DNA had solved the mystery of who killed Jane Durrua, an eighth-grader who was raped, beaten and strangled 36 years earlier.
"Through DNA, we put a face to the killer of Jane Durrua, and that face belongs to Jerry Bellamy," prosecutor John Kaye said.
The killer, however, turned out to be someone else.
Two years after Bellamy's arrest, investigators discovered that evidence from the murder scene had been contaminated by DNA from Bellamy, whose genetic sample was being tested at the same lab in an unrelated case. He was freed. Another man ultimately was arrested in the killing but died before trial.
DNA has proved itself by far the most effective and reliable forensic science. Over the last two decades, it has solved crimes once thought unsolvable, brought elusive murderers and rapists to justice years after their misdeeds and exonerated the innocent. In courtrooms and in the popular imagination, it is often seen as unassailable.
But as the nation rushes to take advantage of DNA's powers, it is becoming clear that genetic sleuthing also has significant limitations:
* Although best known for clearing the wrongfully convicted, DNA evidence has on occasion linked innocent people to crimes. In the lab, it can be contaminated or mislabeled; samples can be switched. In the courtroom, its significance has often been overstated by lawyers or misunderstood by jurors.
* The rush to collect DNA and build databases has in some cases overwhelmed the ability of investigators to process the evidence and follow up on promising leads. Some crime labs have huge backlogs of untested evidence, including thousands of rape evidence kits. In some cases, criminals who could have been caught have offended again.
* Debates have flared over civil rights and privacy, presaging possible constitutional challenges to DNA collection and storage. Critics object, for instance, to storing DNA from people arrested but not convicted of crimes and from suspected illegal immigrants.
In Britain, which has the world's most aggressive approach to forensic DNA, a legal backlash has already begun. The European High Court of Human Rights ruled this month that the country's indefinite storage of DNA from people merely arrested for crimes violated privacy rights. Britain has until March to submit plans for destroying samples or to make a case for keeping them.
In the U.S., authorities are plunging ahead with a dramatic databank expansion.
A California initiative passed in 2004 will permit authorities, starting in January, to store DNA from anyone arrested on suspicion of felonies and serious misdemeanors, even if they are not ultimately convicted.
California's database is expected to swell by about 300,000 DNA profiles next year, bringing the total to 1.4 million.
The FBI's national database, which already contains 6.4 million profiles, is projected to add about 1.3 million annually from federal arrestees and illegal immigrants alone.
Nothing to fear
When the California law, Proposition 69, passed, it was widely believed that the innocent had nothing to fear from having their genetic profiles in a database, said UC Irvine criminology professor William Thompson, considered the U.S. leading authority on DNA laboratory error.
Now, he said, "when you look at all the errors that have come to light around the world -- and we're only finding the tip of the iceberg -- it really raises concerns about how many people you want to have in a database. There are certainly doubts in my mind whether I would want to be in one."
Through the California Public Records Act, The Times obtained documents from five state-run and three county forensic labs reporting scores of laboratory errors or "unexpected" resultsover a five-year period ending in 2007. Labs must track these outcomes and keep them on file under state and federal rules.
Thompson, who reviewed the records for The Times, said that "on a regular basis, laboratory personnel make mistakes that could lead to false identifications" of suspects.
The records show, for instance, that between 2003 and 2007, the Santa Clara County district attorney's crime laboratory caught 14 instances in which evidence samples were contaminated with staff members' DNA, three in which samples were contaminated by an unknown person and six in which DNA from one case contaminated samples from another.
The records also revealed three instances in which DNA samples were accidentally switched, one in which analysts reported incorrect results and three mistakes in computing the statistics used in court to describe the rarity of a DNA profile.
The number reported was small considering overall caseload -- 3,100 over five years -- but Thompson said mistakes caught by labs "undoubtedly" make up a small fraction of errors. (In fact, he said, labs that report the most are probably better run than those that claim none.)
The leading cause of false DNA database matches is cross-contamination of samples, Thompson said.
An incident in a state-run lab in Sacramento illustrated how easily this can happen: DNA discovered on a cigarette matched the profile of a sexual assault victim from another case.
Had the assault victim smoked the cigarette? No. Cross-contamination occurred when the sample from the cigarette was processed close to the victim's vaginal sample.
The risk of DNA contamination has "greatly increased" as scientists have learned how to obtain DNA profiles from a billionth of a gram of genetic material, according to a report last year by the Nuffield Council on Bioethics in London, a group that examines developments in biology and medicine.
"The results may therefore be misleading, and yet they could be presented as powerful evidence in a courtroom. This makes it vital that defendants are not convicted on a DNA match alone," the report said.
Jonathan Jay Koehler, a professor at Arizona State University who has studied lab error, estimated the rate of false DNA matches at about 1 in 1,000, whether they are caught or missed.
"No one would ride on an airline that crashed one out of every 1,000 flights," he said.
Other experts counter that lab errors are rare and will be caught if the proper precautions are taken. "If they are rare enough, then we should be cold-blooded" about isolated injustices, said Charles Brenner, an Oakland-based forensic mathematician.
Jill Spriggs, chief of California's Bureau of Forensic Services, said California's forensic labs are among the best in the nation. "Errors are detected -- if there are any -- based on quality control measures that we have in place."
Wrongful incriminations from DNA evidence have pierced the science's image of infallibility. When Alan Nelson, father of a woman who had been murdered with her daughter in Australia, learned earlier this year that the wrong man had been arrested because of a contamination, he was incredulous: "I thought the DNA was 100% perfect," Nelson told the Herald Sun, an Australian newspaper.
Despite such cases, DNA evidence holds great sway among jurors.
Dan Krane, an Ohio-based DNA expert who has testified in about 75 cases, mostly for the defense, was shaken by the conviction of Gary Leiterman, a Michigan man sentenced in 2005 to life in prison for the nearly four-decade-old murder of a law student.
"It hit me like a ton of bricks," Krane said. "How could that possibly have happened?"
Leiterman, a nurse, came under suspicion when the Michigan state crime lab found a "match" between his genetic profile and evidence from the murder victim's stockings.
But DNA from a blood drop on the victim's hand matched someone else, a man who was 4 at the time of the crime.
It turned out that DNA samples from Leiterman and the other man were being analyzed at the crime lab, as part of separate criminal cases, at about the same time as the murder evidence. Leiterman's DNA was at the lab because of a conviction for prescription fraud. Krane, who testified on Leiterman's behalf, said contamination was the most likely explanation for the findings.
The prosecutor had another explanation: Leiterman killed the student. He was linked to the crime by other evidence, including a handwriting sample and a gun he owned that was similar to the murder weapon.
The prosecutor admitted that the 4-year-old could not have been the killer but speculated that he might also have been at the crime scene, with a nose bleed. A jury convicted Leiterman of first-degree murder. He is appealing the verdict.
Fingering the wrong suspect can be a sign of broader sloppiness in a lab.
A Sacramento crime lab analyst had to resign two years ago after his superiors discovered he had relied on substandard DNA evidence in a rape case to produce a profile that matched an innocent man. The lab later learned the analyst had failed to check his work in dozens of other cases.
But even the most scrupulous analyst can err when interpreting complex DNA "mixtures" -- samples that contain DNA from more than one person -- which turn up more frequently as labs use more sensitive tests.
"If you show 10 colleagues a mixture, you will probably end up with 10 different answers," Peter Gill, then a chief scientist with Britain's Forensic Science Service, said at a symposium in 2005.
In a 2005 study, American laboratories that analyzed the exact same DNA mixture reported widely different statistics on the rarity of the genetic profiles of the contributors.
Estimates of how often the DNA profiles occurred in the general population ranged from 1 in 400,000 people to 1 in 1,000,000,000,000,000.
In the courtroom, say some leaders in the field, the overwhelming numbers presented to jurors convey a near-certainty the science does not justify.
As much as errors can hurt the wrongfully accused, victims don't see justice served when DNA evidence goes untested.
The nation's labs have a serious backlog of DNA crime-scene samples awaiting analysis, as many as 400,000 cases involving rapes and other violent crimes, according to Human Rights Watch.
Meanwhile, criminals are free to strike again: In Ohio, a man committed 13 more rapes while his DNA sat inside a rape kit in storage for two years.
Earlier this year the Los Angeles Police Department acknowledged a backlog of as many as 7,000 DNA crime samples. The Los Angeles County Sheriff's Department later admitted to having 4,727 untested sexual assault evidence kits.
Law enforcement officials blamed a lack of money and manpower. But the federal government contends that the LAPD and other agencies aren't spending all the money they allocate for backlog reduction. For that very reason, it cut LAPD's $1 million share of the funding in half this year.
"We have the science, we have the technology, we have the capacity to prevent rapes. And we are not using it to its fullest potential," said Gail Abarbanel, director of the Rape Treatment Center at Santa Monica-UCLA Medical Center.
Even when crime scene evidence is analyzed and a suspect identified, authorities don't always pursue conviction.
Nationally, the FBI reported hits in the national database aided 77,791 investigations this year, but no one tracks how many result in convictions.
"Hundreds of DNA database hits languish without any follow up by law enforcement or prosecutors," Frederick Bieber, a Harvard Medical School professor, reported in the journal American Society of Law, Medicine & Ethics, in 2006.
In June 2004, a DNA profile from a sexual assault in Oakland matched that of a convicted child molester. By the time police contacted the man six months later, he had molested another child.
Rockne Harmon, a former prosecutor who is a DNA consultant to the California Department of Justice, is working on a cold-hit tracking system that includes sending detectives e-mails when suspects are identified.
"I don't find it very acceptable that we dumped all those millions of dollars [into the DNA database] and what do we get for our money?" Harmon said. "You could say we get a lot of hits, but what contribution have they made to making society a better place? I am not sure that anybody can really tell you that."
Civil libertarians and other critics, meanwhile, fear that pursuit of DNA's scientific potential risks trampling on the rights of the innocent.
"The ACLU opposed these databases from the beginning because we predicted this exactly," said Tania Simoncelli, science advisor to the American Civil Liberties Union. "The database would be established for one purpose and then used for others."
Some legal scholars say that the collection of DNA from arrestees violates the U.S. Constitution because only the convicted have forfeited their rights to privacy.
Among concerns: DNA can be tested over and over again for new information as technology advances. Innocent people theoretically could see their genomes probed without their knowledge or consent.
In Britain, such concerns have brought what Bob Green, a forensic science director in Britain's Home Office, calls a "push back."
Last year, Britain's Nuffield Council on Bioethics complained that too many people were winding up on the British database for petty crimes they committed as children or for arrests that did not produce convictions.
In its ruling earlier this month, Europe's High Court for Human Rights demanded the destruction of an estimated 850,000 DNA samples and profiles of arrestees and expressed astonishment at "the blanket and indiscriminate nature of the power" of the government to collect and retain DNA.
Even Alec Jeffreys, the British geneticist whose discovery led to DNA profiling, is dismayed by the number of juveniles and innocent people now in the database. When Jeffreys made his discovery in 1984, he figured DNA would be a tool of "the last resort" in fighting crime.
"I couldn't have been more wrong," he said.
Times researcher Vicki Gallay contributed to this report.