We implemented a rapid protocol to train teachers in a school for detained juvenile offenders to implement the Good Behavior Game, an empirically supported group contingency procedure, in their classrooms. Data were collected on both teacher (i.e., two measures of treatment integrity) and student behavior (i.e., disruptive behavior). Results indicated clear treatment effects in three of five classrooms and low levels of disruption but unclear effects in the other two. Most students and teachers preferred the intervention and found it effective.