OBJECTIVE - We sought to investigate whether a polymorphism in the alcohol dehydrogenase 1c (ADH1C) gene modifies the association between alcohol consumption and type 2 diabetes. RESEARCH DESIGN AND METHODS - In nested case-control studies of 640 women with incident diabetes and 1,000 control subjects from the Nurses' Health Study and 383 men with incident diabetes and 382 control subjects from the Health Professionals Follow-Up Study, we determined associations between the ADH1C polymorphism, alcohol consumption, and diabetes risk. RESULTS - Moderate to heavy alcohol consumption (>5 g/day for women and >10 g/day for men) was associated with a decreased risk of diabetes among women (odds ratio [OR] 0.45 [95% CI 0.33- 0.63]) but not men (1.08 [0.67-1.75]). ADH1C genotype modified the relation between alcohol consumption and diabetes for women (Pinteraction = 0.02). The number of ADH1C*2 alleles, related to a slower rate of ethanol oxidation, attenuated the lower risk of diabetes among women consuming ≥5 g alcohol/day (Ptrend = 0.002). These results were not significant among men. Results were similar in pooled analyses (Pinteraction = 0.02) with ORs for diabetes among moderate drinkers of 0.44 (95% CI 0.21- 0.94) in ADH1C*1 homozygotes, 0.65 (0.39 -1.06) for heterozygotes, and 0.78 (0.50-1.22) for ADH1C*2 homozygotes compared with those for ADH1C*1 homozygote abstainers (Ptrend = 0.02). CONCLUSIONS - ADH1C genotype modifies the association between alcohol consumption and diabetes. The ADH1C*2 allele, related to a slower oxidation rate, attenuates the lower diabetes risk among moderate to heavy drinkers. This suggests that the association between alcohol consumption and diabetes may be causal but mediated by downstream metabolites such as acetate rather than ethanol itself.