This paper describes the life-style and behavioral correlates of change in coronary heart disease risk factors measured eight years apart in the young adult offspring of the Framingham Heart Study cohort. Changes in total cholesterol, lipoprotein cholesterols (high density lipoprotein (HDL) cholesterol, low density lipoprotein (LDL) cholesterol, very low density lipoprotein (VLDL) cholesterol), and blood pressure were observed longitudinally in 397 men and 497 women who were aged 20-29 years at entry into the study. Stepwise multiple linear regression procedures were used to identify characteristics and their changes that were significantly associated with risk factor changes in each sex. The attribute most strongly and consistently related to lipoprotein and blood pressure changes in both sexes was change in body mass index (p less than or equal to 0.01 or p less than or equal to 0.001). In addition to weight gain, increases in alcohol consumption in men (p less than or equal to 0.001) and beginning oral contraceptive use in women (p less than or equal to 0.01) were associated with increases in blood pressure over the study period. Weight loss, stopping or decreasing cigarette consumption (p less than or equal to 0.01), increasing alcohol intake (p less than or equal to 0.01), and, in women, discontinuing oral contraceptive use (p less than or equal to 0.01) also were independently related to improvements in lipoprotein profiles during follow-up. After adjustment for all life-style correlates of risk factor change, simple self-assessments of physical activity or activity change were negatively associated with changes in VLDL cholesterol (p less than or equal to 0.01) and the total cholesterol/HDL cholesterol ratio (p less than or equal to 0.05) in men and positively associated with changes in HDL cholesterol (p less than or equal to 0.05) in women. Sociodemographic and behavioral characteristics that made a further independent contribution to increases in the total cholesterol/HDL cholesterol ratio in men were blue-collar occupation and trait Type A behavior pattern (p less than or equal to 0.05). Unexplained, but provocative, results of this study included the associations of interim vasectomy with increases in total cholesterol in men (p less than or equal to 0.05) and of number of livebirths with decreases in total cholesterol and HDL cholesterol in women (p less than or equal to 0.01). These findings are among the first to offer prospective evidence which suggests that habits and behaviors during young adulthood have a substantial effect on lipid and lipoprotein profiles in men and women.(ABSTRACT TRUNCATED AT 400 WORDS)