We showed that wind and dust information and incidence in the early dry season predict part of the year-to-year variability of the seasonal incidence of meningitis at both national and district levels in Niger. Models of this form could provide an early-season alert that wind, dust, and other conditions are potentially conducive to an epidemic.