A treatment package consisting of a DRO procedure, token fines, and prompted relaxation was used to reduce the agitated-disruptive behavior of a person with autism and mental retardation living in a community group home. The agitated-disruptive behaviors (cursing, hitting, kicking, throwing objects, and verbal threats) were measured during three different activities in a group home. The baseline rates of the agitated-disruptive behavior during one of these activities was relatively low, during another was moderate, and during the third was high. DRO procedures were partially implemented by a peer with Down syndrome and mental retardation during a portion of the study. Effects of the DRO procedures were as follows: During each activity an initial reduction of agitated-disruptive behavior was dependent on choosing an appropriate DRO interval, with shorter DRO intervals required during activities in which the baseline rates of the agitated-disruptive behavior were higher. Once shorter DRO intervals had been used to reduce agitated-disruptive behavior, longer DRO intervals were effective in maintaining those reductions. Reductions were maintained for up to 6 months.