Dawson et al.’s findings led us to suspect that children with ASD would evidence more pronounced difficulties relative to control sample (CS) children on audiovisual processing tasks involving ‘‘human’’ audio and visual stimuli (e.g. faces and voices) than tasks involving ‘‘nonhuman’’ stimuli (e.g. bouncing balls). The present study further explores differences in the use of visual information between children with and without ASD, and it begins to address whether children with ASD perform typically on similar kinds of tasks with nonhuman stimuli. A McGurk task (McGurk and MacDonald 1976) was used to assess audiovisual processing along with several control conditions that were designed to determine whether children with ASD could process audio and visual information if the task involved an explicit comparison of the two modalities. The study also included several different types of audiovisual mismatches to determine whether the children with ASD possessed a general processing difficulty, or whether processing difficulties were more pronounced with stimuli involving human faces and voices than with stimuli involving objects (e.g. bouncing balls).