We measured the detection of motion before, during and after a saccade to explore the effects of a saccade on motion perception. To isolate the low-level motion mechanism, the stimulus was a random-dot field displaced by small distance (0.3 deg) within a stationary frame. The displacement signaled motion clearly if eyes were fixated, but for the displacement during a saccade, motion was not detected whether the displacement was defined in spatial coordinates (expt 1) or in retinal coordinates (expt 2). Since motion could be seen with ISIs longer than the duration of a saccade (expt 3), the suppression cannot be attributed to visual loss during the saccade. Experiment 3 also showed that motion was never seen for a displacement that occurred during a saccade, even though the random dots were replaced by a uniform field during the eye movement thereby eliminating any masking effect of the sweep of the image across the retina. The purpose of the saccadic suppression of motion may be to block out unreliable motion signals that would be produced by a saccade. Since saccade distances are very often greater than the maximum distance over which the low-level motion mechanism can produce accurate direction discrimination for fine textures, motion signals would generally indicate false directions if they were not suppressed.