We evaluated contingent and noncontingent practice of an alternative behavior (appropriate play) as intervention for stereotypy that occurred with two play objects in a 5-year old boy with autism. Contingent practice was implemented as a consequence for stereotypy, and the rate of noncontingent practice was matched (yoked) to that schedule. On average, stereotypy with one object (blocks) was less frequent during implementation of both contingent and noncontingent practice, but neither procedure had a sustained positive effect with the second play object (figures). Appropriate play did not improve consistently with intervention. Issues related to treating object stereotypy are discussed.