We examined the structural and predictive properties of the Psychopathy Checklist-Revised (PCL-R) in large samples of Canadian male Aboriginal and non-Aboriginal offenders. The PCL-R ratings were part of a risk assessment for criminal recidivism, with a mean follow-up of 26 months postrelease. Using multigroup confirmatory factor analysis, we were able to show that the PCL-R items were invariant across these 2 groups and that a 4-factor model fit the data well. Predictive accuracy analyses (receiver operator characteristic curves and Cohen's "d") generated effect sizes that were medium in magnitude overall for the PCL-R total score in the prediction of violent, nonviolent, and general criminal recidivism (area under the curve = 0.63-0.70, Cohen's d = 0.28-0.42) for both ancestral groups. When disaggregated into its constituent factors, for both ancestral groups, the Lifestyle and Antisocial factors consistently and significantly predicted all recidivism outcomes, whereas the Interpersonal and Affective factors did not significantly predict any of the recidivism outcomes. Finally, structural equation modeling results with the total sample indicated that the PCL-R factors were able to account for 32% of the variance in a latent recidivism factor. Implications regarding the latent structure of psychopathy and the clinical use of the instrument with Aboriginal and non-Aboriginal male offenders are discussed. (Contains 3 tables, 2 figures and 4 footnotes.)