We studied the effects of ketamine, an antagonist of the N-methyl-D-aspartate receptors, on (1) the spontaneous saccades, (2) the vestibulo-ocular reflex (VOR), and (3) the optokinetic nystagmus (OKN) in 8 cats. Ketamine was given intramuscularly at four dosages (1, 2, 8, and 16 mg/kg). Eye movements were measured using the magnetic field-search coil technique. Ketamine did not prevent the occurrence of saccades, but each of them was followed by a centripetal postsaccadic drift. The time-constant of the drift induced by ketamine was 1.0 s when the given dosage was 1 mg/kg and 0.35 s when the given dosage was 16 mg/kg. Post-saccadic drift caused by a low dosage of ketamine may reflect only a mismatch between the pulse and the step commands that create saccades. The highest used dosages of ketamine aggravated the post-saccadic drift probably by disturbing the oculomotor neural integrator. To elicit the horizontal VOR, the head was submitted either to sinusoidal rotations (+/- 20 degrees; 0.05 to 1 Hz) or to a rotation at a constant velocity (100 degrees/s during 40 s). In darkness, the VOR step gain was reduced by ketamine in a dosage-dependent manner. VOR phase lead at 0.10 Hz oscillation in darkness increased from 4.0 degrees +/- 2.4 degrees to 51.6 degrees +/- 7.5 degrees after administration of ketamine at 16 mg/kg. This suggests that ketamine, at least at higher dosages, induces a failure of the neural integrator. Chemical blockade of the vestibular commissure by ketamine may also be responsible for the reduction of the VOR gain. Horizontal OKN was tested using a step stimulus (30 degrees/s during 40 s). When ketamine was given at 1 mg/kg, the average steady-state gain of the OKN diminished from 0.6 +/- 0.2 to 0.3 +/- 0.1. After administration of ketamine at 2 mg/kg, the OKN was abolished. The sensitivity of OKN to ketamine is explained at least partly by the fact that ketamine acts against the visual pathways in the retina, in the geniculate nucleus, and in the visual cortex. The time course of the optokinetic afternstagmus (OKAN) and that of the decrease of the prerotatory and postrotatory VOR were not reduced by ketamine administered at 1 or 2 mg/kg. This shows that ketamine does not affect the velocity-storage mechanism at these dosages.