The ability to perceive structure using motion information was examined using a reaction time task with two primate species. Homo sapien and Macaca mulatta subjects were quantitatively tested under identical conditions to detect the change from a control unstructured to a test structured motion stimulus. The structures underlying the test were rotations of a plane, expansion of a plane, and a rotation of a three-dimensional cylinder. On many of the stimulus conditions, the two species performed similarly, although there were some species differences. These differences may be due to the extensive training of the monkeys or the use of different cognitive strategies by the human subjects. These data provide support for the existence of a neural mechanism that uses flow fields to construct two- or three-dimensional surface representations.