1. Cats were trained to track a small target by rewarding them for keeping their eyes on target. Eye movements were measured by the electromagnetic search coil technique. 2. Cat saccades are qualitatively similar to primate saccades, but exhibit more variability in their parameters. However, they have longer durations and lower maximum velocities than primate saccades. As in the monkey, the duration of the horizontal or vertical component of an oblique saccade is lengthened when the orthogonal component has a larger amplitude. Cat saccades can be modified in midflight like human saccades. Opening the visual feed-back loop by controlling target position with eye position causes the cat to execute a staircase of equal amplitude saccades if a retinal error is present. Increasing the amount of visual feed-back induces saccadic oscillations. 3. Horizontal smooth pursuit of a 0 . 5 deg visual target is limited to velocities of less than 1 deg/sec. However, moving an optokinetic background with the 0 . 5 deg target enables the cat to achieve higher horizontal smooth eye velocities of up to 8 . 5 deg/sec. Prolonged (10-20 sec) constant velocity rotation of an optokinetic drum evokes horizontal slow-phase velocities of up to 28 deg/sec. In response to vertical movements of the target and optokinetic background, smooth eye movements reached 6 deg/sec maximum upward velocities but only 2 . 5 deg/sec maximum downward velocities. Opening the feed back loop with no retinal error present causes the eye to exhibit a growing smooth trajectory. The response to a Rashbass step-ramp target suggests that the feline smooth response is a function of target movement rather than displacement. 4. These data suggest that cat saccadic eye movements resemble those of primates while the cat smooth pursuit and optokinetically induced eye movements are more similar to those of the rabbit.