AbstractTo test the hypothesis that the facial nerve conduction latency test is a better and earlier indicator of prognosis than other electrodiagnostic tests, 86 patients with Bell's palsy were followed for a minimum of 4 months. To select control subjects for our own research clinic and for comparison with the patient population, latency values in 25 normal volunteers (50 sides) were determined. Serial maximal stimulation tests (MST) and latency tests were conducted to determine disease severity and prognosis in Bell's palsy patients. Outcome was graded using the Facial Paralysis Recovery Profile (FPRP) and Facial Paralysis Recovery Index (FPRI) as well as the House grading system. The capability of the two tests to accurately predict outcome was evaluated. The MST accurately predicted outcome in 94% of patients studied. In the control group, normal latency values were a mean 3.8 msec with a standard deviation of 0.49. In the patient population, latency values were either within normal limits or absent. When done within 4 days of onset of Bell's palsy, neither test was capable of predicting axonal degeneration. Statistical analyses included Fisher's Exact Test, the paired Student'sttest, and correlation coefficient calculations.