Anecdotal evidence indicates that dyslexia is positively associated with superior visuospatial ability but empirical evidence is inconsistent. We explicitly tested the hypothesis that dyslexia is associated with visuospatial advantage in 20 dyslexic and 21 unimpaired adult readers using paper-and-pencil measures and tests of "everyday" visuospatial ability. We found no main effect of group on visuospatial task performance but did observe a series of significant sex x group interactions. Relative to dyslexic and unimpaired women, dyslexic men were significantly better at identifying shapes in ambiguous figures, reproducing complex figures, reproducing designs using coloured blocks, and recalling the direction of the Queen's head on a postage stamp. They were also significantly faster and more accurate than unimpaired men at navigating and recreating a virtual environment. These data suggest that visuospatial advantage in dyslexia may be confined to men. Possible explanations for this are discussed. (Contains 3 tables and 3 figures.)