A long-latency stretch reflex (LLSR) has been described in the human masseter muscle, but its pathway remains uncertain. To investigate this, the excitability of corticomotoneuronal (CM) cells projecting to masseter motoneurons during the LLSR was assessed with transcranial magnetic stimulation (TMS). A facilitated response to TMS would be evidence of a LLSR pathway that traverses the motor cortex. Surface electromyogram electrodes were placed over the left or right masseter, and subjects ( n=10) bit on bars with their incisor teeth at 10% of maximal electromyographic activity (EMG). Servo-controlled displacements were imposed on the lower jaw to evoke a short- and long-latency stretch reflex in masseter. TMS intensity was just suprathreshold for a response in contralateral masseter. Trials consisted of: (1) stretch alone, (2) TMS alone, and (3) TMS with a preceding conditioning stretch at varied conditioning-testing (C-T) intervals chosen to combine TMS with the short-latency stretch reflex (3 ms, 5 ms) and the LLSR (23-41 ms). Masseter EMG was rectified and averaged. With TMS alone, mean (+/- SE) MEP area above baseline was 56+/-9%. The area of masseter MEPs above baseline in the C-T trials was calculated from each EMG average following subtraction of the response to stretch alone. Conditioning muscle stretch had no significant effect on masseter MEPs evoked by TMS with any C-T interval (ANOVA; P=0.90). In addition, subjects were unable to modify the SLSR or LLSR by voluntary command. It is concluded that the long-latency stretch reflex in the masseter does not involve the motor cortex and is not influenced by "motor set".