The research aimed to aid understanding of the safe and effective management of mentally disordered offenders, because this issue is a paramount concern in decisions for community placement. However, nonforensic sites were chosen for the study under the theory that more accurate data collection would be possible when severe sanctions were not likely to be imposed. The participants included 61 males and 29 females ranging in age from 19 to 75 years. Participants completed structured diagnostic interviews. Staff nurses responsible for medication were used to make estimates of medication compliance. Results revealed that the crucial symptoms associated with medication noncompliance were anger, delusions, and hallucinations. As an initial analysis, a stepwise discriminant analysis was moderately successful at predicting medication noncompliance. Findings suggested that it would be helpful for forensic mental health professionals to use the SADS-C, a 45-item Schedule for Affective Disorders and Schizophrenia, rather than the full schedule used in this study, to assess treatment response, past treatment compliance, and targeted symptoms in offenders with mental disorders. Table and 39 references (Author abstract modified)