The relationship of alcohol intake to plasma lipids and lipoproteins was assessed in 1603 white children, ages 12-19 years, from six Lipid Research Clinics as part of the Lipid Research Clinics Collaborative Population Studies. Of the 1603 children, 933 came from a randomly recalled group and 660 from a group recalled because of elevated cholesterol or triglyceride or both (the hyperlipidemic recall group). Using multiple regression analysis, the relationships of lipoproteins (as dependent variables) to alcohol, smoking, age and body mass (as explanatory variables) are assessed in both recall groups. In the random recall group, high-density lipoprotein (HDL) cholesterol was positively related to alcohol intake, independent of the other variables considered; for every ounce of alcohol intake, HDL cholesterol was 0.55 mg/dl higher in males and 1.04 mg/dl higher in females. HDL cholesterol was strongly and inversely related to smoking and body mass in both males and females and was inversely related to age in males. In females, plasma low-density lipoprotein (LDL) cholesterol, triglycerides and very low density lipoprotein (VLDL) cholesterol were all positively related to alcohol intake. In the hyperlipidemic recall group of children, alcohol intake had a weak positive relationship with HDL cholesterol in males; in the females, for every ounce of alcohol intake, HDL cholesterol was higher by 1.5 mg/dl. Alcohol intake was positively related to triglyceride levels in hypertriglyceridemic male children. In each recall group, alcohol intake had a small, significant, positive association with HDL cholesterol levels in 12--19-year-old children, and a less consistent positive association with triglyceride and VLDL cholesterol. If low HDL cholesterol concentrations in children are undesirable, attention should first be focused reduction of smoking (inversely associated with HDL cholesterol) and weight (inversely associated with HDL cholesterol, positively associated with LDL cholesterol, triglyceride and VLDL cholesterol), as measures that may modify HDL cholesterol levels.