We evaluated the effects of positive practice overcorrection (PP OC) on levels of motor stereotypy and appropriate engagement in the activity practiced during treatment with 3 young men with autism. We also measured preference for the practiced activities during preference probes to determine if these activities might acquire aversive properties as a result of the frequent pairing with PP OC. Treatment reduced motor stereotypy for all 3 participants, and engagement increased for 2 of the 3 participants. Relative preference for the activities was not disrupted by the implementation of PP OC, although overall contact with the activities decreased for 1 participant. Results from 1 participant suggest that PP OC may be less effective when stereotypy results in access to a more highly preferred activity.