Electrophysiological studies (for example, electroneuronography, nerve action potentials, absolute amplitudes of the muscle compound action potentials, and stimulation thresholds) do not accurately predict facial nerve function after the excision of acoustic neuromas. To eliminate individual nerve variability, the authors measured the ratio of the amplitudes of muscle compound action potentials produced by stimulating the facial nerve at the brainstem proximally and at the internal auditory meatus near the transverse crest distally after total tumor excision in 20 patients. The mean tumor size was 36 mm. The facial nerves were anatomically intact in all patients after tumor excision. The follow-up period ranged from 14 to 28 months. Facial nerve outcome was determined by a modified House-Brackmann grading scale. Initial facial nerve function was measured at Days 4 to 7 postoperatively, and final function was the grade at last follow up. The following results were obtained: all patients with proximal-to-distal amplitude ratios greater than 2:3 had Grade III or better initial function and Grade I final facial nerve function; 90% of patients with amplitude ratios between 1:3 and 2:3 had Grade III or worse initial facial nerve function, and 100% of these patients had Grade III or better final facial nerve function; all patients with amplitude ratios less than 1:3 had Grade IV or worse initial and final facial nerve function. The authors conclude that the proximal-to-distal amplitude ratios after acoustic neuroma excision can accurately predict postoperative facial nerve function.