1. Although much is known about the neurons that control saccadic eye movements, the precise manner in which they interact is still uncertain. To test the validity of competing models of the pontine saccade generator, neurotoxic lesions were made in the nucleus raphe interpositus (rip), which contains one of the principal types of saccade-related neurons, the omnipause neurons (OPNs). The correlated changes in eye movement were quantified in three juvenile rhesus macaques and compared with the results predicted by different models. 2. After the location of the OPNs was mapped, the rip was subjected to sequential, punctate pressure injections of ibotenic acid. The resulting progressive damage was correlated with changes in saccade metrics, including a decrease in peak saccadic velocity and an increase in saccade duration. 3. The damage to rip and presumably to the OPNs was not associated with a change in the animals' ability to maintain steady fixation of a stationary target. 4. The results suggest that Robinson's original local feedback model of saccade generation should be modified. Either a second integrator should be added or the concept of local feedback should be abandoned entirely. 5. The suggestion that the OPNs are primarily responsible for fixation is probably incorrect. OPNs may contribute to fixation stability along with a number of other sources.