The visual world is filled with moving objects, and it is therefore no surprise that the visual system contains special mechanisms for the analysis of motion (Hildreth 1983; Nakayama 1985). Visual motion processing has recently been the subject of vigorous neurobiological, psychophysical, and computational analysis. We now have a good understanding of the neuronal mechanisms that represent motion signals in the cerebral cortex (Maunsell and Newsome 1987) and sophisticated models of their function on a number of different levels (Hildreth and Koch 1987). Much interest in visual motion processing mechanisms has come from a desire to link psychophysical studies of motion analysis to their physiological substrate (Newsome et al., this volume). Other investigators have sought to relate computational approaches to motion sensing to biological mechanisms (Adelson and Bergen 1985; Watson and Ahumada 1985). Relatively little attention has been paid to the role that motion signals play in the generation of eye movements. Visual mechanisms that analyze form perform optimally only when the retinal image is stationary or slowly moving (see Graham 1990). One of the main roles of the oculomotor system is to steady the images of moving objects so that they can be properly seen; to perform this stabilization, the oculomotor system must have accurate information about retinal image motion. Signals about motion drive several kinds of eye movements, including the smooth pursuit eye movements that match the velocity of eye movement to the velocity of an attended moving visual target.