Estimating the risk of sexual recidivism for a juvenile sex offender is essential in order to protect public safety by identifying and evaluating high risk adolescents and to ensure the rights and welfare of juvenile offenders who will not likely reoffend. Empirically validated risk assessment methods are needed to aid in the classification and treatment of juvenile sex offenders. The present study utilized a dataset collected by Maricopa County, AZ, and aggregated by the National Juvenile Court Data Archive. The purpose of the study was to evaluate and characterize risk factors for juveniles who have been charged with a sexual offense in order to determine the predictive utility of these factors for subsequent offending, as well as offense trajectory, and to evaluate risk factors for nonsexual offenders who have committed crimes of various severities. The results of the present study show the strongest individual predictors of sexual recidivism to be prior nonsexual offending, prior sexual offending, hands-off offending, offending against a child, younger school grade/age at time of initial offense, Asian or Hispanic ethnicity, and not attending school. A preliminary screening measure was developed from the seven positive risk factors, and ROC analysis produced an AUC indicating moderate predictive utility in discriminating between juvenile sex offenders who would sexually reoffend and those who would not. Copyright © 2013 John Wiley & Sons, Ltd.