A meta-analysis was conducted to examine whether the predictors of recidivism for mentally disordered offenders are different from the predictors for nondisordered offenders. Effect sizes were calculated for 35 predictors of general recidivism and 27 predictors of violent recidivism drawn from 64 unique samples. The results showed that the major predictors of recidivism were the same for mentally disordered offenders as for nondisordered offenders. Criminal history variables were the best predictors, and clinical variables showed the smallest effect sizes. The findings suggest that the risk assessment of mentally disordered offenders can be enhanced with more attention to the social psychological criminological literature and less reliance on models of psychopathology.