ABSTRACT
Risk assessments have been used in the juvenile justice system to predict future crime, identify youth needs, and inform case planning. Currently, the literature shows that juveniles are most often assessed when they enter court jurisdiction and this entry or initial risk assessment score is typically used to predict recidivism. The current study used the Youth Level of Service/Case Management Inventory to examine the predictive validity of initial, exit, and change in risk scores on a diverse sample (N = 360) of young probationers. This study is one of few attempts to determine the predictive accuracy of change in risk, and the first to investigate the comparative validity of initial and exit risk scores with a sample of young offenders. Exit and change in risk scores significantly predicted recidivism 1-year post probation, yet initial scores did not. Differential predictive validity across race and gender was not identified for the initial or exit risk scores. The importance of examining changes in risk over time was discussed.