OBJECTIVES: This study examined the association between intake of whole vs refined grain and the risk of type 2 diabetes mellitus. METHODS: We used a food frequency questionnaire for repeated dietary assessments to prospectively evaluate the relation between whole-grain intake and the risk of diabetes mellitus in a cohort of 75,521 women aged 38 to 63 years without a previous diagnosis of diabetes or cardiovascular disease in 1984. RESULTS: During the 10-year follow-up, we confirmed 1879 incident cases of diabetes mellitus. When the highest and the lowest quintiles of intake were compared, the age and energy-adjusted relative risks were 0.62 (95% confidence interval [CI] = 0.53, 0.71, P trend < .0001) for whole grain, 1.31 (95% CI = 1.12, 1.53, P trend = .0003) for refined grain, and 1.57 (95% CI = 1.36, 1.82, P trend < .0001) for the ratio of refined- to whole-grain intake. These findings remained significant in multivariate analyses. The findings were most evident for women with a body mass index greater than 25 and were not entirely explained by dietary fiber, magnesium, and vitamin E. CONCLUSIONS: These findings suggest that substituting whole- for refined-grain products may decrease the risk of diabetes mellitus.