Although children initially identified as having elevated levels of coronary heart disease risk factors tend to have elevated follow-up levels (track), a substantial proportion show regression to the mean. Relationships of both genetic and environmental factors to tracking of serum lipids and lipoproteins over an 8-year period were examined in children from a biracial community. Associations between initial and follow-up levels were similar at 3, 5, and 8 years of follow-up; however, differences between the race-sex groups were observed. Two initial measurements reduced the number of nontrackers (children with high initial and decreased follow-up levels). Fathers of children showing persistently elevated levels of either serum total cholesterol or triglycerides were more likely to have diabetes mellitus or to have had heart attacks than were fathers of nontrackers. Children tracking for elevated low-density lipoprotein cholesterol levels had larger increases in triceps skinfold thickness than did the nontrackers. Children consuming alcohol tended to have elevated levels of high-density and decreased levels of low-density lipoprotein cholesterol at follow-up. These observations indicate that the use of repeated serum lipid and lipoprotein determinations, along with measurements of obesity and information concerning family history of heart attack and diabetes mellitus, can aid in the prediction of future elevated serum lipid and lipoprotein levels.