Correlates of high-density lipoprotein (HDL) cholesterol and other lipids and lipoproteins were studied in white men ages 40-59 who were part of the 15% random sample recalled to Visit 2 of the Lipid Research Clinics Program Prevalence Study. Standardized examinations were conducted by two U.S.S.R. and nine U.S. clinics. Mean plasma lipid and lipoprotein cholesterol levels differed significantly between the two countries, with the U.S.S.R. subpopulations having higher mean total plasma and HDL cholesterol levels and HDL/total cholesterol ratios and lower mean triglyceride levels and low-density lipoprotein (LDL)/HDL cholesterol ratios than the U.S. subpopulations. Small, but statistically significant, differences were found in some dietary components. The U.S.S.R. sample had a significantly higher intake of saturated fatty acids, carbohydrates, complex carbohydrates, and kilocalories/kilogram body weight and a significantly lower intake of total fat, polyunsaturated fatty acids, protein, and polyunsaturated/saturated fat ratio. The multiple regression models tested were not major predictors for total plasma cholesterol or LDL cholesterol. Characteristics associated with higher HDL cholesterol levels in both countries were lean body mass, ethanol consumption, abstinence from cigarette smoking, and lower dietary consumption of carbohydrates.