1. In the ketamine-anesthetized guinea pig, electromyographic (EMG) responses of the digastric muscle and vertical and horizontal movements of the mandible were studied when loci within the caudal pontine and rostral medullary reticular formation were stimulated during rhythmic jaw movements (RJMs) evoked by stimulation of the masticatory area of the cortex. 2. Within these regions electrical brain stem stimulation of the pontis nucleus caudalis and nucleus gigantocellularis (PnC-Gi) of the reticular formation completely blocked RJMs at stimulus intensities as low as 10 microA while suppressing the short-latency digastric EMG response that was time locked to each cortical stimulus in the train. PnC-Gi stimulation did not, however, reduce the excitability of the short-latency corticotrigeminal excitatory pathway to digastric motoneurons when tested by short pulse train stimulation at 2 Hz (3 pulses, 500 Hz, 0.3 ms) in the absence of RJMs. 3. Short trains (80 ms) of PnC-Gi stimuli delivered at various phases of the RJM cycle produced a permanent phase shift of the RJM rhythm. If the stimulus train was delivered at an early phase of the cycle (8-40%) the next cycle onset was advanced; if the train was delivered later in the cycle (60-80%) the next cycle onset was delayed. Long trains of PnC-Gi stimuli (100, 200, 300, and 400 ms) increased the time of onset of the next cycle by an amount directly proportional to the duration of the stimulus train. 4. Digastric EMG activity occurring during cortically evoked RJMs occupied nearly 50% of the cycle. If a short train of PnC-Gi stimuli was delivered between approximately 5 and 125 ms after the onset of the burst, the duration of the burst was significantly shortened. 5. These results demonstrate that the suppression of cortically evoked RJMs resulting from PnC-Gi stimulation is due to direct effects on central circuits responsible for the production of the RJM behavior and not on the motoneurons themselves. The evidence presented is consistent with our previously presented hypothesis that the neurons involved in mediating the short-latency corticotrigeminal pathway to digastric motoneurons are separate and distinct from those neurons comprising the central networks responsible for the production of the fundamental jaw oscillation during RJMs.