the current study examined the predictive validity of two adolescent instruments, using outcome measures of recidivism. the psychopathy checklist:youth version (pcl:yv) was retrospectively coded from file information on 102 young offenders referred to an out-patient setting for multi-disciplinary assessment; the youth level of service/case management inventory (yls/cmi) was previously completed by probation workers on the same sample. a 3% year fbllow-up period occurred prior to examination of criminal records. the predictive accuracy of both instruments was examined using area under the curve (aucs). aucs for general and violent recidivism ranged from .57 to .81, and .57 to .79, respectively. the pcl:yv outperformed the yls/cmi in most areas. implications for follow-up are discussed.