An inverse association between moderate alcohol consumption and coronary artery disease has been demonstrated in epidemiologic studies of diverse design. These include ecologic correlations, case-control, longitudinal and clinical studies. The consistency, strength and independence of the inverse relationship argues persuasively for a causal association. These data also suggest that both abstention and heavy alcohol use are associated with an increased risk for coronary artery disease. The effect of moderate alcohol consumption on lipoprotein and apolipoprotein levels is a biologically plausible and likely mechanism for this inverse association. Alcohol consumption elevates HDL cholesterol, although it is unclear whether the HDL subfractions HDL-2 and HDL-3 are beneficially altered. Recent evidence, however, suggests that the apolipoproteins may be more important indicators of coronary artery disease, and moderate alcohol consumption does beneficially alter these proteins. Alcohol may also affect coronary artery disease by other mechanisms, which may include fibrinolytic activity, coagulation, blood pressure, coronary vasoreactivity, and sociobehavioral factors.