Texas court pauses execution again in shaken baby murder case
The Texas Court of Criminal Appeals on Thursday issued another pause in the execution of Robert Roberson, a man convicted nearly two decades ago in the death of his young daughter. The execution had been scheduled for Oct. 16, marking the third time a date has been set for him since 2016. The pause comes as his case is sent back for further review in the trial court.
State officials cited procedures tied to a 2013 statute that allows condemned prisoners to challenge convictions when key scientific evidence is later found unreliable. The statute is often described in coverage as addressing “junk science,” and it gives courts a pathway to reopen cases where expert testimony no longer meets current scientific standards. Roberson’s legal team invoked that statute to ask for a fresh look at the medical evidence that helped secure his conviction.
At trial in 2003, prosecutors said Roberson struck and violently shook his 2-year-old daughter, Nikki Curtis, causing catastrophic head trauma. The prosecution argued that her injuries were consistent with what was then called shaken baby syndrome, and the jury found Roberson guilty of murder. He was sentenced to death following that conviction.
Roberson has consistently maintained his innocence and spoke to reporters in recent days to deny the allegations. “I never shook her or hit her,” Roberson said, repeating the short, stark declaration he has made for years. His lawyers contend the case rests on now-contested medical conclusions rather than airtight forensic proof.
Medical opinion in the case has split along clear lines. Some experts and family members of Nikki Curtis have stood by the original finding of inflicted trauma and have argued that Roberson had a history of violent behavior toward the child. Other medical experts called during appeals and post-conviction reviews have suggested that the fatal condition may have had alternative causes, including complications from illness such as pneumonia.
The debate over what used to be called shaken baby syndrome is part of a broader scientific shift. Over the last decade, researchers have re-examined the markers once thought to be definitive, and some courts have concluded that older interpretations are unreliable. That evolving science underpins the legal argument that evidence once treated as settled should not be the basis for an irreversible punishment like execution.
Roberson’s attorneys also raise procedural complaints separate from the science. They allege judicial misconduct related to steps taken after Nikki Curtis became critically ill, claiming the presiding judge authorized actions that interfered with Roberson’s parental rights and allowed others to withdraw life support. Those assertions, if proven or if they prompt new factual findings, could alter both the legal narrative and the factual record presented to jurors in 2003.
The case has a long history of stops and starts. A previous execution date was set nearly a year ago before it was also stayed, and records show this is the third time state authorities have calendared Roberson for execution since his conviction was upheld. Each new development has brought renewed attention to the evidence and to the motions filed by the defense that seek to exploit changes in scientific understanding.
Sending the matter back to the trial court means judges there will be asked to reevaluate what evidence was relied upon, how the testimony was framed, and whether new expert opinions change the picture enough to require relief. That work typically includes allowing new experts to explain modern findings, reviewing old expert reports with fresh eyes, and testing the procedural claims now asserted. It is not a quick process and can produce a range of outcomes from denial to a full new trial.
For families on both sides, the pendulum swing of hope and dread is intense. Victims’ relatives who believe the original verdict brought justice may see another review as a painful delay. At the same time, defendants and their advocates argue that the finality of the death penalty demands the highest possible certainty before it is carried out.
Texas remains one of the most active death-penalty states, and courts there are increasingly sorting through cases where forensic science has evolved since trials were held. Judges and appeals courts across the country have grappled with the tension between honoring prior verdicts and acknowledging the responsibility to correct convictions tainted by now-discredited methods. Roberson’s case sits at that crossroads, testing how those principles play out in practice.
Legal observers note the practical steps that typically follow a remand: scheduling hearings, lining up new expert witnesses, and reexamining trial records. The trial court could hold evidentiary hearings where experts testify under oath about current science and where the defense presents its judicial-misconduct allegations in detail. Prosecutors will have a chance to defend the original findings and to argue that the conviction should stand despite new interpretations of the evidence.
Whatever the outcome of the trial court’s review, more appeals are likely, and the case could work its way back up to appellate courts. Each stage will test legal standards about new evidence, the reliability of expert testimony, and whether potential errors were harmless or fatal to the integrity of the verdict. That process could take months or longer, leaving the scheduled execution date in limbo.
At its core, the matter raises a central question about the criminal justice system: how to reconcile changing science with the moral and legal weight of capital punishment. Roberson’s case will likely be watched by defense attorneys, prosecutors, scientists, and families as a measure of how courts balance finality with fairness. For now, the immediate consequence is procedural; the execution is paused and the legal review begins anew.
