Serum lipids, blood pressure and body mass may mediate the U-shaped relationship of alcohol consumption with type 2 diabetes mellitus and coronary heart disease. This study examines the cross-sectional and long-term longitudinal relationships of (changes in) alcohol consumption with (changes in) serum lipids, blood pressure and body mass indices. In this prospective, observational cohort study, two measurements of alcohol consumption, serum total and high-density lipoprotein cholesterol, triglycerides, blood pressure, body weight, the thickness of four skinfolds and waist circumference were performed 4 years apart in healthy volunteers (143 men and 174 women, 32 years old at the first measurement). Alcohol consumption from beer, wine and distilled spirits was assessed using an extensive dietary history interview. Linear regression analyses were performed to study the cross-sectional relationships between the amount of alcohol consumed at the age of 32 years and the levels of the lipids, blood pressure and body weight indices, and to study the longitudinal relationships between the changes in the amount of alcohol consumed over the 4 years of follow-up and the concurrent changes in the lipids, blood pressure and body weight indices. Nonlinearity was investigated for the cross-sectional relationships. A 10-g/day difference in alcohol consumption was positively related with a 0.05 mmol/L (1.9 mg/dl) difference in high-density lipoprotein cholesterol in both cross-sectional (p = .004), and longitudinal (p < .0001) analyses. This relationship did not differ for men and women or for the consumption of beer, wine or distilled spirits. Relationships with changes in total cholesterol, triglycerides, systolic, diastolic, and pulse pressure, body weight and the sum of four skinfolds were not significant. A borderline significant inverse longitudinal relationship was found with waist circumference. The other lifestyle behaviors (tobacco smoking, physical activity and dietary habits) were major confounders of most cross-sectional relationships between alcohol and serum lipids, blood pressure and body mass indices. The longitudinal relationships, however, were not confounded by changes in the other lifestyle behaviors. A significant nonlinear relationship was found for systolic blood pressure, in which drinkers of about 30 g/day had the lowest values. Moderate alcohol consumption and moderate long-term changes in alcohol consumption are positively related with the levels and changes in high-density lipoprotein cholesterol in healthy adult men and women. A moderate inverse association between alcohol and waist circumference may be expected. No relationships were found with triglycerides, blood pressure, body weight and the sum of the thickness of four skinfolds. Other lifestyle behaviors confound the cross-sectional, but not the longitudinal, relationships between alcohol consumption and serum lipids, blood pressure and body mass indices. Gender and type of beverage do not modify the relationships between alcohol consumption and these indices.