School-aged children with autism spectrum disorders (ASDs) have delayed gestural development, in comparison with age-matched typically developing children. In this study, an intervention program taught children with low-functioning ASD gestural comprehension and production using video modelling (VM) by a computer-generated robot animation. Six to 12-year-old children with ASD (N = 20; IQ < 70) were taught to recognize 20 gestures produced by the robot animation (phase I), to imitate these gestures (phase II) and to produce them in appropriate social contexts (phase III). Across the three phases, significant differences were found between the results of the pretest and the immediate and follow-up posttests; the results of both posttests were comparable, after controlling for the children's motor and visual memory skills. The children generalized their acquired gestural skills to a novel setting with a human researcher. These results suggest that VM by a robot animation is effective in teaching children with low-functioning ASD to recognize and produce gestures.