The contribution of a visual source of contextual information to speech perception was measured in 12 listeners with aphasia. The three experimental conditions were: Visual-Only (referential gesture), Auditory-Only (computer-edited speech), and Audio-Visual. In a two-alternative, forced-choice task, subjects indicated which picture had been requested. The stimuli were first validated with listeners without brain damage. The listeners with aphasia were subgrouped as having high or low language comprehension based on standardized test scores. Results showed a significantly larger contribution of gestural information to the responses of the lower-comprehension subgroup. The contribution of gesture was significantly correlated with the amount of ambiguity experienced with the auditory-only information. These results show that as the auditory information becomes more ambiguous, individuals with impaired language comprehension deficits make greater use of the visual information. The results support clinical observations that speech information received without visual context is perceived differently than when received with visual context.