OBJECTIVE 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. RESEARCH DESIGN AND METHODS We followed 57,053 participants of the Danish Diet, Cancer, and Health cohort in the Danish National Diabetes Register between baseline (1993–1997) and 27 June 2006. We estimated the mean levels of nitrogen dioxide (NO2) at the residential addresses of the cohort participants since 1971 and modeled the association between NO2 and diabetes incidence with a Cox regression model, separately for two definitions of diabetes: all cases and a more strict definition where unconfirmed cases were excluded. RESULTS Over a mean follow-up of 9.7 years of 51,818 eligible subjects, there were 4,040 (7.8%) incident diabetes cases in total and 2,877 (5.5%) with confirmed diagnoses. Air pollution was not associated with all diabetes cases (hazard ratio 1.00 [95% CI 0.97–1.04] per interquartile range of 4.9 μg/m3 mean NO2 levels since 1971), but a borderline statistically significant association was detected with confirmed cases of diabetes (1.04 [1.00–1.08]). Among confirmed diabetes cases, effects were significantly enhanced in nonsmokers (1.12 [1.05–1.20]) and physically active people (1.10 [1.03–1.16]). CONCLUSIONS Long-term exposure to traffic-related air pollution may contribute to the development of diabetes, especially in individuals with a healthy lifestyle, nonsmokers, and physically active individuals.