The current project was an initial attempt to develop assessment procedures for distinguishing between obsessive-compulsive (OC) and stereotypic behavior and evaluate the impact of different treatments for these behaviors. Two individuals with autism, one with repetitive behavior characteristic of OC behavior and one with repetitive behavior not characteristic of OC behavior, participated in the study. In Experiment 1, given that individuals with Obsessive-Compulsive Disorder (OCD) report experiencing unpleasant urges that are relieved when they perform compulsive actions, an attempt was made to identify these experiences by measuring heart rate and affect when access to repetitive behavior was restricted and allowed. In Experiment 2, a multiple schedules treatment was conducted with each participant, and in Experiment 3, the participant with autism and OC behavior completed exposure and response prevention (ERP) treatment. The overall results across studies suggest that one potential way to discriminate between OC behavior and stereotypy in nonvocal children with autism is to consider the topography of repetitive behavior along with changes in physiology and affect. In addition, it may be worth considering the use of ERP, a traditional treatment for OCD, to treat repetitive behavior maintained by automatic reinforcement if treatments that provide access to repetitive behavior are not effective.