Optimal strategies for identifying children with hypercholesterolemia have not been established. Several groups have advocated that testing of serum cholesterol levels be limited to those children who have family histories of hyperlipidemia or premature coronary heart disease. We studied the ability of comprehensive family histories to identify children with hyperlipidemia in a group of 114 children (mean age, 8 +/- 4 years) who were referred for treatment of hypercholesterolemia. A positive family history was defined according to guidelines of the American Academy of Pediatrics. The mean fasting total cholesterol in the children was 5.74 +/- 1.42 mmol/L (222 mg/dL). Family history was negative for hypercholesterolemia or premature coronary heart disease in 22 (22%) of 100 children with total cholesterol levels greater than the 75th percentile for their ages, in 13 (18.3%) of 71 children with total cholesterol levels greater than the 95th percentile for their ages, and in four (11.8%) of 34 children with presumed heterozygous familial hypercholesterolemia. Of the 78 children who had both hypercholesterolemia and positive family histories, hyperlipidemia was reported in 72 families, whereas premature heart disease was reported in only 27. We conclude that in a population of children referred because of known hypercholesterolemia, a detailed family history not only fails to identify many children with mild hypercholesterolemia, but also fails to identify a significant proportion of children with markedly elevated cholesterol levels. Additionally, in families of children with hypercholesterolemia, a history of hyperlipidemia is more common than a history of premature heart disease.