Flores v. Texas
Brief filed: 3/26
Court: Supreme Court of the United States
Year of Decision: 2026
Issue
Whether psychological research on eyewitness memory demonstrates that initial non-identifications are highly probative of innocence and that subsequent in-court identifications, made after repeated exposure to a suspect, suggestive procedures, and the passage of time, are inherently unreliable.
Index Topic
Eyewitness Identification Research
Facts
This is a capital case involving the reliability of eyewitness identification. On January 29, 1998, Jill Barganier witnessed two white males with long hair exit a car in her neighbor's driveway and enter the house through the garage; her neighbor was found murdered later that morning. Barganier described both men as white males with long brown hair and helped create a composite sketch of the car's driver, which was used to identify Richard Childs, who later confessed to the shooting. Six days after the murder, Barganier underwent hypnosis conducted by a patrol officer performing his first hypnosis, during which she again described the passenger as having long, dark hair. The hypnotist asked leading questions and told Barganier she might remember more later. Barganier then created a digital composite sketch depicting a white male with shoulder-length hair, but police showed her a photo array containing six Hispanic males with very short or shaved hair, including petitioner Charles Don Flores. Barganier did not recognize anyone in the array. Over the following months, images of petitioner appeared in news reports, at least one of which Barganier saw. Approximately 13 months after the murder, Barganier identified petitioner for the first time at trial. The jury did not hear evidence of Barganier's initial description of the passenger as a white male with long hair, her composite sketch consistent with that description, or her failure to recognize Flores in the photo array. Despite the absence of physical evidence linking him to the murder, petitioner was convicted and sentenced to death.
APA's Position
A substantial body of psychological research on human memory and eyewitness identification demonstrates that the eyewitness identification at the heart of this capital case is deeply unreliable. The APA’s brief addresses three key findings from the research. First, initial non-identifications by eyewitnesses are highly probative of innocence. Eyewitness identification relies on recognition memory, which is automatic and does not require a deliberate search of memory; if a witness does not recognize a person during an initial identification procedure, it is unlikely the witness would later correctly identify that same person. Studies consistently show that lineup rejections are diagnostic of the suspect's innocence, and an emerging expert consensus confirms that initial non-identifications provide more reliable evidence of innocence than previously believed. Second, eyewitness identifications made after repeated exposure to a suspect, suggestive investigative procedures (including leading questions and hypnosis), and the passage of time are likely unreliable. Repeated exposure can cause "unconscious transference," where a witness attributes a suspect's familiarity to the crime rather than to a prior viewing, and suggestive procedures can create confident but false memories. Third, in-court identifications are inherently suggestive and highly prone to inaccuracy. The courtroom setting makes it obvious which person is the defendant, witnesses feel pressure to identify someone, and the same factors that undermine reliability also artificially inflate the witness's confidence, making such identifications both unreliable and powerfully persuasive to juries. As the largest professional association of psychologists in the United States, the APA urges the Court to grant certiorari and be guided by this extensive body of psychological research in evaluating the reliability of the eyewitness identification upon which petitioner's conviction and death sentence primarily rested.
Results
On June 15, 2026, the Supreme Court denied the petition for a writ of certiorari. The denial left in place the Texas Court of Criminal Appeals' October 9, 2025 order dismissing Flores's subsequent habeas application without reviewing the merits of his claims.