Individuals with autism spectrum disorder (ASD) have qualitative impairments in social interaction and often prefer food or tangible reinforcement to social reinforcement. Thus, therapists who work with children with ASD often use food or tangible items as reinforcers to increase appropriate behaviors or decrease problem behaviors. The goal of the present study was to shift children's preferences from a highly preferred tangible item to an initially nonpreferred social reinforcer using an observational conditioning procedure. Participants observed a known peer engage in a simple task and select the social reinforcer that was not preferred by the participant. This procedure resulted in a shift of preference toward the social reinforcer by all participants. Maintenance data showed that although the preference change did not maintain for 1 of the participants, it was quickly reestablished with additional observational trials. Results provide further support for the use of observational procedures to alter preferences.