Responses of single neurons were recorded in the medial and descending vestibular nuclei (MVN and DVN) and in the deep cerebellar nuclei of three juvenile rhesus monkeys (Macaca mulatta). Neuronal activity was measured during both passive sinusoidal and nonsinusoidal whole body rotation (peak velocities were under 90 degrees/s) and during active head movements. Although the active head movements occasionally exceeded 300 degrees/s, most exhibited peak velocities of less than 200 degrees/s. A total of 133 units sensitive to horizontal head rotation were recorded, and of these, 38 were held for sufficient time to obtain both passive and active head movement data. Comparison of the neuronal firing patterns obtained during active and passive head movements revealed no apparent differences. Thus neurons that were observed to burst or pause during saccades with the head fixed continued to do so when the head was free. Both the sensitivity to head velocity and the "inferred" spontaneous firing rate were compared during active and passive head movements by plotting rate-velocity curves for both conditions. When the data points were fitted with linear regression lines, no statistically significant differences in either sensitivity or spontaneous rate were found. The present study provides no evidence that efferent vestibular activity alters the properties of afferent vestibular neurons during active head movements, as has previously been suggested (21). Furthermore, neurons in the rostral portions of the vestibular nuclei in primates encode head velocity based entirely on labyrinthine information. Neither neck proprioceptors nor an efference copy of the head movement motor program seem to contribute significantly to the firing patterns observed.