We examined the effects of family history of coronary artery disease (CAD), apolipoprotein E (apo E) phenotype, and lipoprotein(a) [Lp(a)] on the response of plasma lipids to change in dietary lipid intake after 3 mo of nutrition education in 125 children aged 4-10 y. The subjects were healthy children with elevated low-density-lipoprotein (LDL)-cholesterol concentrations who participated in the Children's Health Project, a nutrition-education program designed to lower plasma cholesterol by means of dietary modifications in accordance with recommendations of the National Cholesterol Education Program. Dietary and plasma lipids were measured by three 24-h recalls and assessments of two fasting plasma samples collected before and 3 mo after the start of intervention. Family history of CAD was determined by questionnaires administered to parents at baseline. Apo E phenotyping was done with isoelectric focusing followed by immunostaining; Lp(a) was measured with two-site immunoradiometric assays of frozen aliquots of plasma samples collected at baseline and 3 mo. After adjustment for intervention group, age, sex, and body mass index, analysis of covariance showed that baseline plasma lipid concentrations were the strongest independent predictors of change in plasma lipids after 3 mo. Plasma total and LDL-cholesterol concentrations in children with less family history of CAD were significantly more responsive to change in dietary cholesterol than concentrations in children with a stronger family history of CAD. Neither apo E phenotype nor Lp(a) significantly influenced change in plasma lipids independently or interactively with change in dietary lipids.