OBJECTIVE: A J-shaped association has been demonstrated between alcohol consumption and atherosclerosis. Insulin resistance, also a risk factor for atherosclerosis, has been shown to have a similar J-shaped association with alcohol intake. This raises the question of whether insulin sensitivity (S(I)) is a causal intermediate in the alcohol-atherosclerosis relationship. RESEARCH DESIGN AND METHODS: The Insulin Resistance Atherosclerosis Study was a multicenter cohort study designed to investigate relationships among S(I), risk factors for cardiovascular disease, and carotid artery atherosclerosis. Using regression analysis, we tested whether adjustment for S(I) attenuated the alcohol-atherosclerosis relationship observed at baseline. RESULTS: A J-shaped association was observed between alcohol consumption and common carotid artery intimal medial thickness. The protective aspect of the alcohol-atherosclerosis relationship was attenuated by 25% after the adjustment for S(I). However, an interaction was observed between alcohol consumption and glucose tolerance (GT) status. In comparison with never drinkers, all levels of alcohol consumption were associated with less atherosclerosis in participants with normal GT status. Participants with impaired GT status (but not diabetes) demonstrated a J-shaped alcohol-atherosclerosis association. All levels of alcohol consumption were associated with more atherosclerosis in participants with diabetes. CONCLUSIONS: S(I) may be a causal intermediate at protective levels of alcohol intake, but an alcohol-GT interaction precluded a definitive conclusion. Moderate alcohol consumption may increase the risk of atherosclerosis in people with diabetes. These findings contrast with previous reports and do not support current recommendations regarding moderate alcohol consumption in people with diabetes. More research is needed to clarify this issue.