There is a discrepancy between the averaged amount of retinal slip measured during head movement and the expected level of visual performance. This study offers an explanation based on the sinusoidal velocity profile of the retinal image motion during head oscillation. Contrast thresholds were measured for stationary and moving 0.5 and 6 cycles per degree (c.deg-1) gratings with both passive oscillation (2-6 Hz) and fixation of the head. Eye and head movements were recorded with a magnetic induction coil technique. When expressed in terms of averaged retinal image motion, threshold contrast for the 0.5 c.deg-1 grating remained suboptimal, whereas the blurring of the 6 c.deg-1 grating appeared to be less during the head movement. The data obtained from head movement versus non-head movement conditions could, however, be matched when the oscillatory character of the retinal image motion was taken in to account. The amplitude of gaze movement was small relative to the size of the individual cycles in the 0.5 c.deg-1 grating. This reduced spatiotemporal interaction which rendered their visibility sub-optimal. Contrast thresholds for the 6 c.deg-1 grating were highly correlated to the duration of the time intervals in which the retinal slip was below the critical value of 2 deg.s-1. Intermittent exposure to quasi-stable retinal images explained the relatively high level of contrast sensitivity for the 6 c.deg-1 grating during head oscillations.