A complete description of the loss of contrast sensitivity at high spatial frequencies requires an estimate of the role of eye movements, which could blur fine detail. We describe a new technique to isolate their effect. Observers viewed either a 100 c/deg interference fringe, which the cone mosaic aliased to a low frequency zebra stripe, or an artificial zebra stripe. The real and artificial zebra stripes have similar spatial patterns, but differ in the temporal modulation produced by eye movements. Contrast threshold was measured as a function of duration for both stimuli flashed in the dark. The ratio of the contrast thresholds for the real and artificial zebra stripes with long durations, when eye movements could have a differential effect, is always within a factor of two or so of the ratio for 1 msec flashes, when eye movements are eliminated. These results support the view that eye movements are only a minor source of image degradation even at very high spatial frequencies, and provide no support for the view that they improve high resolution tasks.