This study tested the hypothesis that children with autism are impaired in using verbal encoding and rehearsal strategies in the service of working memory. Participants were 24 high-ability, school-age children with autism and a comparison group matched on verbal and non-verbal IQ, receptive and expressive vocabulary, and visual memory. Working memory was assessed using verbal and non-verbal variants of a non-spatial, self-ordered pointing test [Petrides, M., & Milner, B. (1982). Deficits on subject-ordered tasks after frontal- and temporal-lobe lesions in man. Neuropsychologia, 20, 249-262] in which children had to point to a new stimulus in a set upon each presentation without repeating a previous choice. In the verbal condition, the stimuli were pictures of concrete, nameable objects, whereas in the non-verbal condition, the stimuli were not easily named or verbally encoded. Participants were also administered a verbal span task to assess non-executive verbal rehearsal skills. Although the two groups were equivalent in verbal rehearsal skills, the autism group performed significantly less well in the verbal, but not the non-verbal, self-ordered pointing test. These findings suggested that children with autism are deficient in the use of verbal mediation strategies to maintain and monitor goal-related information in working memory. The findings are discussed in terms of possible autistic impairments in episodic memory as well as working memory.