Research has shown that video and live modeling are both effective in teaching new skills to children with autism. An adapted alternating treatments design was used to compare the effectiveness and efficiency of video and live modeling in teaching response chains to three children with autism. Each child was taught two chained skills; one skill was planned to be taught by video modeling and the other was planned to be taught by live modeling. Results showed that two children were able to acquire the response chains assigned to them successfully with both procedures. For the third child, live modeling was found to be more effective than video modeling. When the efficiency of two procedures is considered, no dramatic differences were observed in favor of either modeling across participants.