The clinical application of transcranial magnetic stimulation (TMS) for prognostic evaluation of peripheral facial palsy is discussed. Because the recovery of motor nerve function is determined by the degree of wallerian degeneration and degenerated nerve fibers are not responsive to electrical stimulation, examinations using electrical stimulation are important for the accurate prognostic evaluation of peripheral facial palsy. However, when the facial nerve is stimulated peripheral to the injury site, the prognosis cannot be estimated correctly until the wallerian degeneration is complete at the site of stimulation. Thus, the results of the nerve excitability test (NET) and electroneurography (ENoG) are not significant for the prognostic evaluation until 7-10 days after the onset of the palsy. When the coil of the magnetic stimulator is placed on the temporal portion at a point 6cm above and 3cm behind the external auditory meatus, the facial nerve can be stimulated around the fundus of the internal auditory meatus, which is the central site of the injury in most cases of facial palsy. In normal subjects, the magnetically evoked compound muscle action potential (CMAP) recorded from the electrodes placed on the orbicularis oris muscle is similar to the CMAP in ENoG. But the latency of the CMAP evoked by TMS delays about 1 msec compairing with the CMAP in ENoG. Although the CMAP evoked by TMS can be recorded in about a half of the patients and the amplitude is very low in the early stages of palsy, 83% of the patients showing a response evoked by TMS in the early stages show good prognosis. Patients with a good prognosis can thus be identified using TMS even in the early stages of palsy. The sensitivities and specificities of the examinations used for prognostic evaluation, such as the scoring system for facial movements, NET, stapedial reflex and TMS, are also discussed.