A Murder Conviction Torn Apart by a Bullet

An Inexact Match

The only other evidence linking Kulbicki's gun to the murder came from the FBI lab. For more than three decades, bureau experts had testified that they could tie bullets or bullet fragments from crime scenes to suspects by comparing the lead content to bullets in an ammunition box or in a gun recovered from a suspect.

In 2005, the FBI abruptly stopped using the technique after studies -- including one by the National Academy of Sciences -- found that FBI witnesses had inappropriately suggested to jurors that they could match bullets to specific boxes or guns.

But back in 1995 when Kulbicki was convicted, the science was still in wide use.

Then-FBI examiner Ernest Roger Peele told jurors that the composition of the bullet fragment found in Nueslein's head matched that of the fragment found in Kulbicki's truck. Tests showed that the two fragments "matched at each and every element," Peele testified.

But when Drouet summoned new experts, including the FBI's retired chief metallurgist, she found that the fragments did not match exactly. Different quantities of one of the trace elements -- "arsenic 1" -- were found in the two fragments. Drouet accused prosecutors of ignoring the evidence. "Scientists may not pick or choose among test results," she told the court. FBI officials said that their scientists would sometimes use a second measurement known as "arsenic 2" to compare bullets when the first arsenic measure did not match.

Peele told jurors that the remaining bullets in Kulbicki's revolver did not match the fragments at the crime scene, but that one was close in composition. Prosecutors seized on the remark, suggesting to jurors that the bullets were "very nearly identical" and that this was proof of guilt.

"Out of the billions of bullets in the world, is this just a coincidence that that bullet ended up in the defendant's off-duty weapon?" a prosecutor asked.

In an interview, Peele declined to address Kulbicki or any other specific cases he worked, saying that he may be summoned to testify in appeals. But Peele said his bullet-lead analyses were never intended to be the sole scientific evidence against a defendant.

"It was part of the puzzle. It was part of the situation. It certainly was not a yes-or-no part of the situation. It was not to the point that it made the -- made the case 100 percent," Peele said.

During the trial, prosecutors told jurors that the bullet fragments Kopera and Peele analyzed were "a significant piece of evidence" and a "major link" to proving Kulbicki's guilt, court records show. But now the prosecution suggests that the evidence was not so crucial.

"The evidence from the ballistics examination and comparative bullet analysis was a small portion of the state's case," prosecutors argued in recent motions.

The effort to minimize the importance of the bullet matching comes after the Maryland Court of Appeals in 2006 reversed the murder conviction of Gemar Clemons on grounds that the FBI's bullet-lead science was not based on "generally accepted" scientific principles.

A DNA Factor

The prosecution still counters with one powerful piece of evidence. It says its tests show that the tiny bone fragments recovered from Kulbicki's truck contained DNA from Nueslein and almost certainly came from her skull when the gun was fired.

The analysis of the fragments, however, also is in dispute. One rather large human bone fragment was recovered from the truck after Kulbicki's arrest, but tests for the victim's DNA were inconclusive, Drouet said.

In between Kulbicki's two murder trials, the prosecution's experts tested three much smaller bone fragments that had been sitting for two years in vacuum cleaner bags amid other evidence gathered from Kulbicki's truck. It was those fragments that contained Nueslein's DNA.

During the second trial, the bone-fragment evidence was hard for Kulbicki to overcome. The prosecution expert testified that the DNA in the bone chips matched that of the victim on four of seven measurements and that the likelihood of bone chips matching another person was 1 in 640.

Once again, though, the lab notes Drouet obtained in recent months raised questions about the fragments and the science that was used to test them.

The 1995 notes from the private lab that tested the three smaller bone fragments for the prosecution say that the fragments were "believed to be contaminated."

Despite those concerns, the fragments were not sterilized before DNA testing, Drouet said.

The notes also say that the lab was told not to preserve a small portion of the fragments during the DNA testing, which might have allowed the defense to do its own testing. "OK not to save 10% of sample," the lab notes quote the state as saying. "Do what you can to get results."

It is unclear how the trial judge will rule on Kulbicki. But the case has already exposed a much larger question resonating throughout courts nationwide.

Clifford Spiegelman, a statistician at Texas A&M University who served on the 2004 National Academy of Sciences panel that sharply criticized the FBI's bullet-lead technique, was asked by the defense to review the case. Spiegelman said it mirrors others in which juries relied on prosecution scientists whose testimony is now considered overstated.

"What we're seeing is too many instances in which FBI or other prosecution scientists are simply doing what it takes to 'get their man,' " he said.