The study data provided a reanalysis of data reported by Steadman and others in 1984. The participants were four cohorts of 400 males each, including a prison inmate cohort and a mentally ill patient cohort admitted to institutions in 1968 and 1978. All participants were followed for 11 years from their time of admission. The research focused on age at admission, number of incarcerations, race, mental health hospitalization history, arrest history, violent crimes, and subsequent crimes. Consistent with previous research, results revealed that having a prior arrest history is a stronger predictor of subsequent arrest than is a history of State mental hospitalization. Prisoners with no prior hospitalization consistently posed the highest risk. Some of the data also indicated that mentally disordered inmates pose a slightly lower risk than other inmates, but the difference was not statistically significant. Findings indicated that the presence of a prior arrest was the best predictor of a future arrest, and mental patients with no prior arrests were no more likely and possibly less likely to be arrested than people in the general population. Tables, discussion of study limitations, and 22 references