Numerous efforts are underway to train clinicians in evidence-based practices. Unfortunately, the field has few practical measures of therapist adherence and skill with which to judge the success of these training and implementation efforts. One possible assessment method is using behavioral rehearsal, or role-play, as an analogue for therapist in-session behavior. The current study describes aspects of reliability, validity and utility of a behavioral role-play assessment developed to evaluate therapist adherence and skill in implementing Trauma-Focused Cognitive-Behavioral Therapy (TF-CBT). TF-CBT role-play assessments were conducted with a sample of 43 therapists as part of a larger training study. The TF-CBT role-play assessments were independently coded for TF-CBT adherence and skill by a certified TF-CBT trainer and three clinical psychology doctoral students. Findings indicated good interrater reliability for the individual items (ICC: M = 0.71, SD = 0.15). Regarding utility, 67.19% (n = 43/64) of contacted therapists completed the role-play assessment, which took an average of 30 min (M = 31.42, SD = 5.65) to complete and 60 min (M = 62.84, SD = 11.31) to code. Therapists with a master's degree were more likely to complete the role-play assessment than those with other degrees but no other differences in demographic variables, practice characteristics, or TF-CBT knowledge or training were found between participants and nonparticipants. Role-play assessments may offer an alternative to observational coding for assessing therapist adherence and skill, particularly in contexts where session recordings are not feasible.