The inferior alveolar branch of the trigeminal nerve is at risk during sagittal split osteotomies, a common maxillofacial surgical procedure. The purpose of this study was to evaluate the use of somatosensory evoked potentials to assess the functional state of the nerve during surgery. Ten patients scheduled for bilateral sagittal split osteotomies were studied. Recording electrodes were laced on the scalp overlying vertex and inion, and subcutaneous stimulating electrodes were inserted 1 cm apart over the mental foramina. Recordings were made bilaterally from all patients. Baseline recordings were made after anesthetic induction, prior to the initial incision. Subsequent recordings were made just prior to the mandibular bone cuts, during splitting of the mandible, and after rigid fixation of the mandible. Analysis o f the baseline data identified a series of peaks at 20 and 26 milliseconds following stimulation. This biphasic response was consistent across subjects and was chosen for subsequent analysis. Both the amplitude and latency of the response were significantly (P < .05) affected when the nerve was retracted medially while the medial horizontal bone cuts wre made. However, in all patients, the waveforms returned to baseline values within 10 to 20 minutes. No consistent changes were found in the evoked potentials recorded at other times during the surgical procedure. The results indicate that surgical retraction can cause a transient neurapraxia, but they do not account for postsurgical loss of sensation, which could be due to other factors, such as edema and swelling.