Challenges to verbal encoding may affect the ability of autistic individuals to express their ideas. Therefore, visuospatial expression may represent a person's knowledge and skills more accurately than spoken language. To test this hypothesis, we asked seven autistic adults to linguistically retell and visuospatially reenact several animated clips. On average, visuospatial responses contained more correct elements than spoken responses. The level of intention of the three stimulus categories did not systematically affect response accuracy. Participants who produced visuospatial responses before spoken responses and those who had watched a greater number of stimuli assigned higher intentionality to shapes in the animations that were designed to elicit mentalizing. The modality used for expression may affect accuracy of responses by autistic individuals.