OBJECTIVE: To evaluate the relation between alcohol intake and incidence of type 2 diabetes. RESEARCH DESIGN AND METHODS: This prospective study included 8,663 men with fasting plasma glucose measurements from at least two medical examinations. Alcohol intake was classified into five groups: nondrinkers and four quartiles (Qs) of drinkers according to the amount of alcohol intake. Type 2 diabetes was diagnosed by 1997 American Diabetes Association criteria. RESULTS: There were 149 incident cases of type 2 diabetes during 52,588 person-years of follow-up. There was a U-shaped association between alcohol intake and diabetes, with the lowest incidence of diabetes at Q2 (61.9-122.7 g/week). As compared with Q2, men in Q3 and Q4 had a 2.2- (95% CI 1.2-3.9, P = 0.01) and 2.4-fold (1.4-4.4, P<0.01) risk of developing diabetes, while nondrinkers and men in Q1 had 1.8- (1.0-3.3, P<0.05) and 1.4-fold (0.7-2.6, P = 0.34) higher risk of diabetes, respectively. These associations persisted after adjustment for age, fasting plasma glucose, smoking, BMI, blood pressure, serum triglyceride concentration, cardiorespiratory fitness, HDL cholesterol, waist circumference, and parental diabetes. CONCLUSIONS: We observed an elevated risk of developing type 2 diabetes in nondrinkers and men with high alcohol intakes, when compared with men who reported moderate alcohol intake. Men with a high alcohol intake may be able to reduce their risk of developing type 2 diabetes if they drink less.