Executive functions in 111 adolescent male offenders were examined using the Cognitive Assessment System (CAS). As hypothesized, mean scores for the sample were significantly lower than norms on CAS measures of Planning, Attention, and Successive Processing, and deficits remained for individuals with a childhood-onset when the sample was divided by age-of-onset for antisocial behaviors. Contrary to our hypothesis, mean scores for individuals with adolescent-onset antisocial behaviors were significantly below norms for Planning and Successive Processing. The study provides a direct rather than putative assessment of executive functions and identifies specific executive function deficits regardless of age-of-onset of antisocial behaviors. Implications for practice and limitations of the study are reviewed. (Contains 3 tables.)