1. We report the complex-spike responses of two groups of Purkinje cells (P-cells). The cell were classified according to their simple-spike firing during smooth eye movements evoked by visual and vestibular stimuli with the use of established criteria (Lisberger and Fuchs 1978; Stone and Lisberger 1990). During pursuit with the head fixed, ipsi gaze-velocity P-cells (GVP-cells) showed increased simple-spike firing when gaze moved toward the side of the recording, whereas down GVP-cells showed increased simple-spike firing when gaze moved downward. 2. During pursuit of sinusoidal target motion, the complex-spike firing rate was modulated out-of-phase with the simple-spike firing rate. Ipsi GVP-cells showed increased complex-spike firing during pursuit away from the side of the recording, and down GVP-cells showed increased complex-spike firing during upward pursuit. The strength of the complex-spike response increased as a function of the frequency of sinusoidal target motion. 3. GVP-cells showed directionally selective complex-spike responses during the initiation of pursuit to ramp target motion. Ipsi GVP-cells had increased complex-spike firing 100 ms after the onset of contralaterally directed target motion and decreased complex-spike activity after the onset of ipsilaterally directed target motion. Down GVP-cells had increased complex-spike firing 100 ms after the onset of upward target motion and decreased firing after the onset of downward target motion. As during sinusoidal target motion, each cell's simple- and complex-spike responses had the opposite directional preferences. 4. When the monkeys fixated a stationary target during a transient vestibular stimulus, the retinal slip caused by the 14-ms latency of the vestibuloocular reflex (VOR) affected the complex-spike firing rate. For ipsi GVP-cells, ipsilateral head motion caused transient contralateral image motion and an increase in complex-spike firing. The same vestibular stimulus in darkness caused an almost identical eye movement but had no effect on complex-spike firing. We conclude that complex spikes in ipsi GVP-cells are driven by contralaterally directed image motion. 5. To determine the events surrounding complex-spike firing during pursuit, we triggered averages of eye and target velocity on the occurrence of complex spikes during pursuit of sine-wave target motion. The averages revealed a transient pulse of retinal image motion that peaked approximately 100 ms before the complex spike. We conclude that complex spikes during steady-state pursuit are driven by the retinal slip associated with imperfect pursuit.(ABSTRACT TRUNCATED AT 400 WORDS)