Analyses of convergent validity and group assignment using self-report, caregiver-report and interview-based measures of adolescent psychopathy were conducted in a sample of 160 incarcerated adolescents. Results reveal significant convergent validity between caregiver-report measures of adolescent psychopathy, significant convergent validity between self-report measures of adolescent psychopathy and an interviewer rating scale, but not between the caregiver-report measures and their corresponding self-report measures nor between the caregiver-report measures and the interviewer rating scale. Analyses of group assignment were also poorer than expected among all the measures with none evidencing significant agreement with the expert-rated device (Hare Psychopathy Checklist-Youth Version; PCL-YV), the most common forensic instrument used in clinical practice. Part of the poor agreement may be related to the poor psychometric performance of the callous-unemotional subscale of most of these measures and the low response rates from caregivers (N = 35). These findings suggest that the measures do not provide an interchangeable assessment of callous-unemotional traits and suggest that further refinement of the measurement of callous-unemotional traits in youth may be warranted.