Three of the 59 patients were readmitted and redischarged, but each discharge was handled as a separate case in accordance with the procedures used by the hospital. The followup period averaged 73 months and ranged from 5 to 172 months. The analysis focused on whether the index offenses were sexual and nonsexual and whether the first subsequent reconviction involved a serious offense or a minor offense. Results revealed that a sexual index offense, prior imprisonment, and absolute discharge were associated with higher rates of reconviction. A sexual index offense and a sexual element in the index offense were associated with higher rates of reconviction for serious offenses. Prior drug abuse was associated with higher rates of minor reoffending but with lower rates of serious offending. A positive linear relationship existed between the number of previous convictions and the probability of reconviction. Tables and 24 references (Author abstract modified)