Animal and cross-sectional epidemiological studies suggest a link between air pollution and diabetes, whereas the limited prospective data show mixed results. We studied the association between long-term exposure to traffic-related air pollution and incidence of diabetes.