Bell's palsy is a relatively common disease characterized by the sudden onset of unilateral facial paralysis. Using a centralized surveillance system that contains demographic, military assignment, and medical encounter data of US military service members, the authors estimated rates, trends, and demographic correlates of risk of Bell's palsy during a 2-year period. Poisson regression was used to estimate the independent effects of climate, season, and latitude. From October 1997 to September 1999, there were 1,181 incident cases of Bell's palsy among US service members. The crude incidence rate was 42.77 per 100,000 person-years. Incidence rates increased with age and were higher among females, Blacks, Hispanics, married persons, and enlisted service members. Both climate (adjusted rate ratio for arid vs. nonarid climate = 1.34) and season (adjusted rate ratio for cold vs. warm months = 1.31) were independent predictors of risk of Bell's palsy. Latitude was not a statistically significant predictor when demographic, climate, and season effects were taken into account. The results are consistent with hypotheses regarding viral etiologies (e.g., reactivation of herpes simplex) of Bell's palsy.