Abstract In an attempt to reduce the levels of disruptive behavior, six participants in a school for students with behavior disorders were shown edited videotapes depicting themselves appropriately engaged in classroom activities. The participants were first exposed to a training condition in which both appropriate and inappropriate behaviors of a peer were either observed or observed and recorded. In the intervention condition, the participants either self-observed and self-recorded their behaviors from videotape or simply observed themselves on videotape. The results showed idiosyncratic effects across both conditions. Measures of treatment integrity and treatment acceptability were also conducted. The results of this study are compared with similar research investigating self-observation interventions.