A NJ court ruled shaken baby testimony inadmissible. Could it help Robert Roberson?
Roberson’s legal team hopes the decision will be considered in the death row inmate’s quest for a new trial.
Jamie Landers and the Associated Press, November 21, 2025
New Jersey’s highest court ruled this week that expert testimony on shaken baby syndrome is scientifically unreliable and inadmissible in two upcoming trials. The decision comes as the long-held medical diagnosis has come under intense scrutiny, including in Texas, where a prominent case remains in limbo.
The New Jersey Supreme Court determined Thursday that a diagnosis of shaken baby syndrome, which is also known as abusive head trauma, is not generally accepted within the “biomechanical community” and is therefore not “sufficiently reliable” for admission at the trials.
The 6-1 ruling deals with the trials of two men facing charges in separate cases, where the young victims showed symptoms that have come to be associated with shaken baby syndrome.
The justices, using an abbreviation for the syndrome, concluded in their lengthy decision that “there was no test supporting a finding that humans can produce the physical force necessary to cause the symptoms associated with SBS/AHT in a child.”
However, in a strongly worded dissent, Justice Rachel Wainer Apter said the other justices put more weight on the views of individual biomechanical engineers over the “consensus perspective of every major medical society in the world.” She noted that every other U.S. state allows testimony in court on the syndrome and “every other court that has considered the question” has held such evidence as admissible.
According to the Mayo Clinic, shaken baby syndrome is a result of forcefully shaking an infant or a toddler, which can damage or destroy a child’s brain cells and cause permanent brain damage or even death. Symptoms include bleeding around the brain, brain swelling and bleeding in the eyes.
Prosecutors and medical societies say the syndrome is the leading cause of fatal head injuries in children younger than 2 years of age, with more than 1,000 cases reported in the U.S. each year, according to the National Center on Shaken Baby Syndrome.
But defense lawyers and some in the medical and scientific communities argue that shaken baby diagnosis is flawed and has led to wrongful convictions, pointing to overturned convictions or dropped charges in California, Ohio, Massachusetts and Michigan.
It’s a debate that has played out in Texas courtrooms for more than two decades in the case of Robert Roberson III, who was convicted of capital murder in 2003 having been accused of shaking his 2-year-old daughter, Nikki, to death.
In his 23 years on Texas’ death row, Roberson, 59, has narrowly avoided execution three times — each stay following a challenge to the syndrome.
Roberson was last spared on Oct. 9, one week before he was scheduled to be put to death. The Texas Court of Criminal Appeals sent Roberson’s case back to district court in light of a ruling made last year in Dallas County. In that case, Andrew Roark’s conviction was vacated on the basis that the science of shaken baby syndrome had evolved, and that expert witnesses would have given different testimony had he gone to trial in 2024.
Roberson is awaiting next steps as the state pushes back on proceeding with the evidentiary hearing that could give him a chance to retry his case. Gretchen Sween, Roberson’s lead attorney, told The Dallas Morning News Friday that like Roark, she hopes New Jersey’s decision will “be considered in assessing Robert Roberson’s quest for a new trial.”
New Jersey’s public defender’s office has already hailed the ruling a “landmark” moment, explaining it reflected the importance of relying on “reliable, well-supported scientific evidence” in criminal cases.
“Where the science is uncertain, the stakes are simply too high to permit unsupported expert opinions to decide a person’s guilt or to justify separating children from their parents,” Cody Mason, a managing attorney in the office, said in a statement.