The study employed a matched group, quasiexperimental design. Treated offenders were compared with untreated offenders on variables most consistently related to recidivism (age, criminal history, and index offense). Treated subjects included 176 patients who spent at least 2 years in the therapeutic community program during the period of its most active operation. Hare's Revised Psychopathy Checklist was used to identify psychopaths and nonpsychopaths in order to examine interrelationships among treatment, psychopathy, and recidivism. In almost all cases, comparison subjects had been convicted of some offense and had served prison sentences. Outcome measures were criminal and violent recidivism, and the average followup period exceeded 10 years. Results showed that, compared to no program (in most cases prison), treatment was associated with lower recidivism, especially violent recidivism, for nonpsychopaths and higher violent recidivism for psychopaths. Given the pervasive pessimism about whether psychopaths can change in any significant way, the finding that participation in the treatment program changed the rate of violent recidivism, albeit for the worse, was remarkable. Overall, results strongly supported the clinical and research utility of Hare's Psychopathy Checklist. Reasons why the therapeutic community had different effects on psychopaths and nonpsychopaths are examined. 75 references and 3 tables