This study evaluated the effects of priming and contingent attention procedures on play variability in a child with autism. During baseline, numbers of novel play episodes, different play episodes, and actions occurred at low rates. Priming procedures did not produce desired change. When contingent attention was implemented, significant increases occurred in novel play episodes, different actions, and different play episodes. These results show that attention contingent on variable play episodes can increase the number of novel responses to play materials. The results are discussed within the context of treatment and future research.