This study was a meta-analysis of three assessment measures commonly used in risk assessment with young offenders: youth adaptations of the Level of Service Inventory (YLS/CMI, LSI-SK), the Psychopathy Checklist (PCL-YV), and the Structured Assessment of Violence Risk for Youth (SAVRY). In all, 49 studies were initially identified for possible inclusion, and 44 were used for the primary analyses. All three measures significantly predicted general, nonviolent, and violent recidivism with comparable degrees of accuracy. The substantive findings did not support the superior predictive accuracy of one measure over another for any particular recidivism outcome. The relatively short mean follow-up time for the young offender studies is also noteworthy, especially when compared to the adult literature in which the mean follow-up time in meta-analytic investigations is about 5 years. The weakest predictive accuracy was observed with respect to sexual recidivism for all three instruments, although this was not unexpected given that none of these measures was specifically designed to assess risk for sexual violence. The findings highlight that the ultimate purpose of risk assessment should be the prevention as opposed to the prediction of criminal recidivism. Data were collected from 49 suitable published and unpublished studies representing 8,746 youth who were identified and evaluated for inclusion. Tables, notes, and references