Over the summer, the Texas Forensic Science Commission, which sets standards for physical evidence in state courts, came to an unsettling conclusion: There was something wrong with how state labs were analyzing DNA evidence.

It seemed the labs were using an outdated protocol for calculating the probability of DNA matches in "mixtures"; that is, crime scene samples that contain genetic material from several people. It may have affected thousands of cases going back to 1999.

At first, they assumed the update wouldn't make a big difference — just a refinement of the numbers.

But when a state lab reran the analysis of a DNA match from a murder case about to go to trial in Galveston, Texas, it discovered the numbers changed quite a bit.

Under the old protocol, says defense lawyer Roberto Torres, DNA from the crime scene was matched to his client with a certainty of more than a million to one. That is, you'd have to go through more than a million people to find somebody else who'd match the sample. But when the lab did the analysis again with the new protocol, things looked very different.

"When they retested it, the likelihood that it could be someone else was, I think, one in 30-something, one in 40. So it was a significant probability that it could be someone else," Torres says.

The change didn't affect the outcome of that case because there was other evidence against his client, but officials in Texas have just begun the process of correcting the mistake.

"We have to go back and identify which of those cases involved DNA mixtures where the lab may have given incorrect results," says Jack Roady, the district attorney in Galveston. "It's going to be a herculean task, but we're gonna do it."

Roady has been cooperating with the Texas Forensic Science Commission on fixing the problem, and he recalls the scene in September when he described the situation to a meeting of his fellow prosecutors.

"There was sometimes moments of collective gasps," he says. "The fact that this science may not have been done correctly in the past gives us great pause."

It's unsettling to find out DNA analysis can vary like this because it threatens to undermine the deep faith people have placed in the technology.

"And it's not faith they should not have had to begin with," says Keith Inman, who teaches forensic science at California State University, East Bay.

Inman has worked with DNA evidence since the 1980s. He says forensic DNA-matching is based on sound science, but sometimes labs can get ahead of themselves. What happened in Texas, he says, is that labs have been using cutting-edge "testing kits" that can extract tiny traces of DNA from crime scenes, but those samples were then analyzed with math that's not suited to "weak" samples that combine DNA from many people.

He says the problem isn't limited to Texas. He says the newest, best analysis method — called "probabilistic genotyping" — takes time to roll out, and that's put labs in a quandary.

"There's this interim time that cases are coming up and the analyst has to do something with it, and they know by definition that there is a better approach," Inman says.

Meanwhile, the justice system's hunger for DNA evidence just keeps growing. There are now police departments that have made swabbing for DNA part of their routine.

"We collect DNA evidence daily," says Jim Ferraris, deputy chief in Salem, Ore. His department has taken advantage of quicker testing provided by the state lab, and he says every officer in town is now trained to collect DNA. They even swab stolen cars and burgled homes.