There are several effective training packages (e.g., behavioral skills training, video modeling, and self-instruction packages) available to train staff. Despite their efficacy, these training procedures require substantial time or preplanning and resources to create materials. Teach-back, an empirically validated method used in the healthcare setting to enhance communication between clinicians and patients, does not require any preplanning or materials. However, this method has yet to be investigated in the context of training and supervision. The purpose of this experiment was to evaluate the efficacy of teach-back in training participants to implement preference assessments and a token economy. The teach-back method improved procedural integrity to at least 88%, and the addition of vocal-verbal feedback resulted in all participants achieving 100% integrity in all skills. We discuss the implications of these findings.