To determine the extent to which cholesterol measured in childhood is predictive of values in adulthood, the investigators conducting the second follow-up of the Beaver County Lipid Study tracked the cholesterol values of 295 adults who had initially participated as children (ages 11 to 14 years) in a countywide school screening program. The follow-up study was conducted 16 years after the initial study, when the participants had reached a mean age of 28 years. The overall correlation (r) between baseline (1972-1973) total cholesterol values and the values found at the present follow-up was 0.44 (p less than 0.0001). Women had a higher correlation (r = 0.51) than men (r = 0.38). In addition, the efficacy of childhood screening for cholesterol levels was assessed by considering currently recommended borderline values (greater than 175 mg/dl (4.6 mmol/L) for children and greater than 200 mg/dl (5.2 mmol/L) for adults) as a "positive" test result. The sensitivity of screening at age 12 years for predicting elevated adult total cholesterol concentrations was 63%, specificity was 67%, and the predictive value of a positive test result was 47%. Comparison of false-positive results (above the borderline cutoff point as a child but not as an adult) and false-negative results (below the borderline cutoff point as a child but above it as an adult) showed that male subjects with false-positive results smoked significantly less than those with false-negative results (p less than 0.05) and had a greater improvement during the preceding 7 years in cholesterol-lowering dietary practices (p less than 0.01). Female subjects with false-positive results smoked significantly less than those with false-negative results (p less than 0.05), were less overweight (p less than 0.05), and had a lower prevalence of oral contraceptive use (p less than 0.01). These results support the potential value of screening for hypercholesterolemia in childhood on a population basis. Although some subjects were misclassified as a result of childhood screening, some of this misclassification was associated with adopting changes that a screening and intervention program would be designed to promote--nonsmoking, weight control, and a prudent diet.