Results of this longitudinal study demonstrated that rapists and mixed offenders had higher psychopathy scores than did the child molesters and incest offenders. The Psychopathy Checklist-Revised (PCL-R) was found to be a weak predictor of sexual recidivism but consistently predicted nonsexual violent recidivism and general recidivism. Sexual deviance measured by a structural rating scheme predicted sexual recidivism. Sexual deviance was a stronger predictor of sexual recidivism than psychopathy. However, the two interacted significantly suggesting that psychopathy could potentiate sexual recidivism. Although psychopathy was a strong predictor of general nonsexual recidivism, sexual deviance was inversely related and no interaction was observed between psychopathy, sexual deviance, and nonsexual recidivism. The relationship of psychopathy to sexual deviance and their link to sexual recidivism has generated considerable interest. This article examined the base rate of psychopathy in different sex offender types using different PCL-R cutoffs, and as a function of the relative contributions of PCL-R Factors 1 and 2. It also investigated the predictive accuracy of PCL-R total, Factor 1 and 2 scores for sexual and nonsexual recidivism using multiple recidivism criteria among groups composed of different sex offender types. Lastly, it examined a possible interaction between PCL-R ratings of psychopathy and sexual deviance in relation to sexual and general recidivism. Participants included 156 male Federal sex offenders admitted between 1983 and 1997 to the Clearwater Sex Offender Treatment Program in Canada. Tables, figure, and references