Gunshot Residue Evidence in a Massachusetts Murder Trial: Commonwealth v. Williams

By Attorney Joe Serpa | Georgetown University Law Center

October 19, 2026

On September 22, 2026, the Supreme Judicial Court held that a judge may not let a jury hear an opinion that gunshot residue particles were found on a defendant as evidence that the defendant fired a gun, unless the Commonwealth first proves that the particle count is reliable evidence of having fired a shot. The jury in the case heard that police found one particle on the defendant’s hand and one on his T-shirt, even though the State Police counted both tests as negative. The Commonwealth’s own expert testified that nothing could be drawn from the results. The court vacated the defendant’s first-degree murder conviction and ordered a new trial.

A Suffolk County jury had convicted Keith Williams in 2017 of murder in the first degree, armed assault with intent to murder, assault and battery by means of a dangerous weapon, and carrying a firearm without a license. The charges arose from a shooting at a parade in Dorchester in August 2014. The Supreme Judicial Court vacated all four convictions. The Commonwealth may retry Williams for murder in the first degree. The decision is Commonwealth v. Williams, SJC-12702 (Mass. Sept. 22, 2026).

The Shooting on Blue Hill Avenue

On the morning of August 23, 2014, a crowd gathered on Blue Hill Avenue in Dorchester to watch a parade. Security cameras recorded two groups of young men with a tense encounter at a convenience store. Shortly afterward, a man in the second group pulled a handgun and pointed it toward the first group. A man in the first group then fired at least five bullets toward him. None of the bullets hit the man with the gun or his companion. One bullet killed a woman crossing the street, and another struck a woman sitting on the median in the leg.

No witness identified Williams as the shooter. A witness who lived nearby watched Williams and two companions walk away from the scene, heard one companion tell him to “get rid of it,” and saw Williams dispose of a revolver at a house across the street. Police found a .357 revolver under the porch, and ballistic testing linked it to both victims. Tests of the revolver for DNA and fingerprints were inconclusive. Police stopped and handcuffed Williams and the two companions a short distance away.

The defense at trial was that Williams was not the shooter. Witnesses described the shooter differently, and some of the details did not fit Williams. One defense witness described a shooter whose clothing matched one of the two companions.

The Gunshot Residue Tests

Police collected samples from the hands of all three men at 11:53 A.M., more than three and one half hours after the shooting. Police collected Williams’s T-shirt the next day, and a criminalist collected samples from it thirteen months later. A State Police criminalist examined each sample with a scanning electron microscope. A gunshot residue particle contains three components that are not generally found together in nature. Those components are barium nitrate, lead styphnate, and antimony sulfide.

The State Police and the Federal Bureau of Investigation require three or more particles for a positive test, and a laboratory cannot add up particles found on separate samples. Every test in the case was negative. The criminalist found one particle on one sample from Williams’s hands and one particle on the T-shirt. He found no particles on either companion. The criminalist agreed on cross-examination that there was “nothing to be drawn from” the negative results. He also agreed that the threshold exists in part to account for contamination.

The criminalist testified that gunshot residue particles are resilient and transferable, and that they persist on handcuffs and on surfaces in police cruisers and police stations. Police had handcuffed and transported all three men before collecting the samples.

The Trial Judge Admitted the Particle Counts

Before trial, the Commonwealth asked to present the expert’s testimony that one particle was found on each of the two samples from Williams. The Commonwealth acknowledged that the expert would testify that the results were negative. The Commonwealth cited Commonwealth v. Johnson, 463 Mass. 95, 108-109 (2012), and planned to argue that the jury could infer from the particles that Williams was the shooter.

Williams opposed the request. He argued that the Commonwealth was using a negative result as if it were positive, and that the particles could have come from contamination. He submitted the laboratory report on the T-shirt, which stated that no conclusions could be drawn from the results. He also submitted a Boston police policy stating that it was not yet possible to assess scientifically the significance of a gunshot residue identification. The judge allowed the testimony, ruled that Williams’s objections went to the weight of the evidence, and left it to the jury to decide what inferences to draw.

The Commonwealth then relied on the particles in its opening statement and its closing argument. In closing, the prosecutor told the jury that the three-particle threshold was “not [their] protocol” but a State Police standard, and that the jury could interpret the evidence however it wanted.

The Reliability Holding

The Supreme Judicial Court held that the judge should not have admitted the testimony as evidence that Williams was the shooter. Expert testimony is admissible under the Daubert-Lanigan standard only when the party offering it proves that it is reliable. The Commonwealth had to prove that the method is reliable and that the expert applied it reliably to the facts of this case. Commonwealth v. Hinds, 487 Mass. 212, 220 (2021); Commonwealth v. Davis, 487 Mass. 448, 453 (2021).

The court held that the Commonwealth offered no evidence that detecting one particle in each of two negative results reliably showed that Williams fired the gun. The only evidence on the point was its own expert’s statement that nothing could be drawn from the results. The Commonwealth also offered no measure of reliability for the particle counts. The Commonwealth argued that the jury could draw the inference and that Williams’s objections went to weight, but the court held that the Commonwealth had to lay the foundation first.

The Court Limits Johnson

The Commonwealth relied on Johnson, in which the Supreme Judicial Court said that the jury could infer a greater likelihood of guilt from two particles found on a sweatshirt. The court explained that in Johnson the defendant did not object to the testimony, and that the court had not decided whether particle counts below the positive threshold are reliable evidence that a person fired a gun. The court disavowed any suggestion in Johnson that such counts support the inference by themselves.

The court did not hold that a below-threshold particle count is always inadmissible. Admissibility depends on the foundation the Commonwealth lays and the inference it asks the jury to draw. The court contrasted Commonwealth v. Linton, 483 Mass. 227, 241-242 (2019), in which an expert showed that a DNA result below a minimum threshold still provided relevant data. The Commonwealth in Williams made no comparable showing.

The Error Required a New Trial

The Supreme Judicial Court held that the error was prejudicial. The prosecutor used the particles in both the opening statement and the closing argument. The prosecutor also used them to answer the defense argument that contamination explained the particles. Both companions had also been handcuffed and transported, yet no particles were found on them.

The Commonwealth’s other evidence of identity was not overwhelming. Williams was seen with the companions seconds after the shooting and disposing of the revolver, and one witness described the shooter as the largest of the three men. However, no witness or forensic test identified him as the shooter. The strongest description for the Commonwealth also included details that did not fit him. Another witness described a shooter who matched a companion’s appearance. The court could not conclude that the error had little or no effect on the jury, so it ordered a new trial on the murder, armed assault, and assault and battery charges.

Retrial for First-Degree Murder Remains Possible

Williams also argued that the evidence did not prove deliberate premeditation, which would have barred a retrial for first-degree murder. The court rejected the argument. Deliberate premeditation can form in a few seconds, and the jury could infer from the events that Williams decided to kill the man with the gun before he fired at least five rounds toward him. The Commonwealth may therefore retry him for first-degree murder.

The court separately vacated the conviction for carrying a firearm without a license. The judge had not instructed the jury that the Commonwealth must prove the defendant had no valid license. Commonwealth v. Carleton, 497 Mass. 11, 13 (2026).

Gunshot Residue Evidence After Williams

When the Commonwealth offers gunshot residue evidence that falls short of a positive result, the defense can ask the judge to exclude it unless the Commonwealth proves, before the jury hears it, that it reliably shows what the Commonwealth says it shows. The defense can ask for a hearing under the Daubert-Lanigan standard, often with a defense expert, and can use the laboratory’s own reports, its protocols, and the expert’s own statements about the threshold.

The decision also shows what a defense lawyer can ask in these cases. When did the police collect the samples? Where was the person held and transported before collection? Did the other people tested have the same contact with handcuffs and cruisers? The expert in Williams agreed that particles persist on those surfaces. The answers can support a motion to exclude the evidence and can support cross-examination if the judge admits it.

Serpa Law Office defends firearms charges in the Boston Municipal Court, the District Court, and the Superior Court. Gunshot residue testing is one of several kinds of forensic evidence that can arise in a firearms case.

Key Takeaways

• In Commonwealth v. Williams, SJC-12702 (Mass. Sept. 22, 2026), the Supreme Judicial Court vacated a first-degree murder conviction because the jury heard gunshot residue particle counts from negative test results.
• The Commonwealth must prove that expert testimony is reliable before the jury hears it, including that the method was reliably applied to the facts of the case.
• A laboratory’s own positive threshold, and its expert’s statement that nothing can be drawn from a negative result, can support a motion to exclude particle counts below the threshold.
• The court limited Commonwealth v. Johnson, 463 Mass. 95 (2012), and held that below-threshold counts are not automatically admissible as evidence that a person fired a gun.
• The court did not bar retrial for first-degree murder, and it separately vacated the license conviction under Commonwealth v. Carleton, 497 Mass. 11, 13 (2026).

Related Serpa Law Office resources

• Massachusetts Criminal Defense of Firearms Offenses
• Carrying a Firearm Without a License in Massachusetts: G.L. c. 269, § 10(a)
• When Police May Frisk a Person for a Gun in Massachusetts: Commonwealth v. Vihaan V.
• How I Use Experts in a Massachusetts Criminal Case
• Medical Expert Testimony in Massachusetts Criminal Cases: The Hsu Decision on Abusive Head Trauma
• Cross-Examination in Massachusetts Criminal Trials
• Massachusetts Criminal Charges Defense