Adolescents, and most recently, adolescent females, have emerged as an important population in violence risk assessment and have sparked a debate regarding the downward and gendered extension of the Psychopathy Checklist: Youth Version (PCL:YV). This article evaluates the differential prediction of the three and four-factor models of the PCL:YV for male (n=201) and female (n=55) juvenile offenders using a prospective four and one-half year follow-up (M=3 years) study. Both models of the PCL:YV were significant predictors for boys; however, contrary to findings from studies using shorter follow-up periods, the predictive power was due primarily to the behavioral features of psychopathy. The PCL:YV was not a significant predictor of non-violent or violent recidivism for girls. This study does not lend support for the use of the PCL:YV as a risk factor for girl offenders. More research is needed to understand the application of the psychopathy construct in youth, particularly in girls.