An inverse association between low to moderate alcohol consumption and coronary heart disease has been demonstrated in epidemiologic studies of diverse design. An attempt was made to determine if this association might be due to an effect of alcohol on apolipoproteins A-I and B and to determine if low-dose alcohol intake might have a potentially protective effect by this mechanism in persons at increased risk for coronary heart disease. To address this, an eight-week prospective randomized clinical trial of abstention versus low-dose alcohol consumption, defined as one beverage per day, was conducted in white men, aged 21 to 60 years, most of whom were patients of a preventive cardiology program. Apolipoprotein A-I levels had a mean increase of 9 mg/dl in the 28 participants who drank alcohol compared with a mean decline of 5 mg/dl in the 28 participants who abstained (p less than 0.005). This association was independent of other cardiovascular risk factors. Low-density lipoprotein (LDL)-B levels had a mean increase of 7 mg/dl in both arms of the trial (NS). However, the ratio of apolipoprotein A-I to LDL-B increased by 4 percent in the drinkers and decreased 10 percent in the abstainers (p less than 0.03). No significant changes in mean levels of total high-density lipoprotein (HDL)-, HDL2-, or HDL3-cholesterol were observed with this low dose of alcohol. This effect on apolipoprotein A-I suggests a possible mechanism by which low-dose alcohol may lower the risk of coronary heart disease.