The present study attempted to determine the impact of alternative verdict choices on the decisions of mock jurors. Subjects used in this study as mock jurors were all college undergraduates. They were shown one of two versions of a videotaped simulated murder trial. Both films presented a defendant who appeared to be suffering emotional difficulties, but in one film the defendant had clearly committed the act while in the other film the defendant's actions were less certain. Subjects than gave their individual verdicts and, after deliberation with other subjects, a total jury verdict. The verdicts available to the subjects varied across three conditions such that the subjects in one condition were only allowed to find the defendant to be innocent or guilty. In another condition the subjects could find the defendant innocent, guilty or not guilty by reason of insanity (NGRI). In the third condition the subjects were allowed to choose between innocent, guilty, NGRI, and guilty but mentally ill (GBMI). The results indicated that the addition of the “mental health” verdicts had a significant impact on the decisions of the jurors. In particular, it appears that only defendants who would otherwise have been found innocent were likely to be found NGRI. This study also indicated that the GBMI verdict is very attractive to mock jurors. Indeed, even innocent defendants were found to be GBMI, a form of guilt, when this alternative was made available. These findings raise potentially important constitutional and practical issues for the trial of emotionally disturbed criminal defendants.