Adolescent sexual abusers are a heterogeneous group of offenders that often receive generic assessment and treatment services that are modeled on research findings from adult sex offender samples. The emotional Stroop task has been used to measure deviant sexual interest in adult samples. The purpose of the present study was to test whether the emotional Stroop task could also be used to assess deviant sexual interest in adolescent samples. Three groups of adolescents (a) sexual abusers (n = 24); (b) offending controls (n = 21); and (c) nonoffending controls (n = 21) completed two emotional Stroop tasks related to deviant sexual interest and tests of executive function. Adolescent sexual abusers were significantly slower to color-name some word stimuli than both adolescent offending controls and adolescent nonoffending controls. However, the task was unable to differentiate between the groups on most of the Stroop word categories. Very little research has been conducted with adolescent offender samples and the emotional Stroop task. Reaction time (RT) and Stroop bias outcome data for adolescent samples appear to be more unsystematic and weaker than has been observed in previous adult data. Based on potential difficulties with reading and development, the emotional Stroop task may not be a task suitable for measuring deviant sexual interest in adolescent samples.