Results indicated that the SAQ maintained satisfactory predictive validity on cross-validation with a high-risk correctional psychiatric sample. The SAQ is a 72-item, self-report measure designed to predict recidivism, supervision violations, and related behavior among non-mentally disordered Canadian Federal offenders. Offenders who had been in-patients at the Regional Treatment Center (Ontario) RTC(O) had committed more sex offenses. Longer sentences being served by the RTC(O) would seem to imply a greater severity of individual offenses, given that the samples did not differ in terms of the number of offenses. Data did not allow a meaningful analysis of all factors potentially contributing to the longer sentences in the RTC(O) sample, but it is reasonable to view the longer sentences as a marker of risk. The indication is that SAQ-outcome relationships cross-validated adequately with a high-risk psychiatric correctional sample. Subjects were male offender populations of a number of the institutions, representing a range of security levels, in the Ontario region of the Correctional Services of Canada. The general correctional sample was composed of 42 offenders reasonably characterized as typical of the Ontario Region Federal incarcerated population. Tables, references