Ocular movements of alert cats were recorded by classical electronystagmography techniques during (a) vestibular stimulation (sinusoidal rotation of the cat in complete darkness), (b) optokinetic stimulation (sinusoidal rotation of the visual surroundings around the stationary cat), (c) additive visual-vestibular stimulation (sinusoidal rotation of the cat inside the stationary lighted surroundings), and (d) conflicting visual-vestibular stimulation (sinusoidal rotation of the cat together with the visual surroundings in phase and at the same speed). The stimulus amplitudes and frequencies ranged from 3 to 20 degrees and from 0.025 to 1 Hz, respectively. When tested in darkness, the vestibuloocular reflex (VOR) gain was about 0.9 at 1 Hz. At lower frequencies, this gain was a bit lower and a phase lead was observed. The VOR system was nearly linear. The optokinetic response (OKR) gain was about 1 at lower frequencies but strongly decreased at higher frequencies. A phase lag paralleled that decrease in gain. Furthermore, the smaller the amplitude of the visual stimulus, the better the effectiveness of OKR stabilization. When working in the light, the VOR was in phase with the stimulus and its gain was nearly 1, whatever the frequency and the amplitude. The VOR inhibition was more effective at lower frequencies. In these conditions the system was markedly amplitude-dependent for both gain and phase.