Background: Recent studies in adults suggest that individual dietary fatty acids differ markedly in their effects on serum lipids and lipoprotein levels. However, these associations have rarely been studied in children. Objective: To assess, using regression procedures, the associations in children between specific fatty acids and nonfasting serum lipids and cholesterol after controlling for total energy and total fat intake, Subjects: The sample consisted of 1,182 children who participated in the Child and Adolescent Trial for Cardiovascular Health. The sample was equally distributed across 4 sites (Louisiana, Texas, Minnesota, California). The sample was 48% boys; 71% white, 15% Hispanic-American, 10% African-American, 2% Asian, and 2% from other or unspecified racial/ethnic heritage. Design: In this randomized multicenter trial with 56 intervention and 40 control elementary schools, food record-assisted 24-hour dietary recalls and serum lipid measurements were collected for each child at baseline (3rd grade) and at the 5th grade follow-up. Statistical Analyses: Repeated-measures analysis of variance was used to evaluate the association between nutrient composition of the diet and serum lipids. Independent dietary variables included amount and type of fat, individual fatty acids, protein, carbohydrate, and fiber. The dependent variables were the absolute values of serum total cholesterol (TC) and high-density lipoprotein cholesterol (HDL-C) both at baseline (3rd grade) and at follow-up (5th grade). Results: Increased total fat (b=0.053; P<.03) was associated with increased TC in the model when energy was held constant, whereas increased carbohydrate was associated with decreased TC (b=-0.021, P<.02) and HDL-C (b=-0.010, P<.005) levels. Increased total protein (b=0.017, P<.05) was associated with increased HDL-C when energy was held constant. Saturated fat (b=0.004, P<.04), unsaturated fat (b=0.004, P<.03), and myristic fatty acid (b=0.021, P<.01) all increased TC in the model when total fat and total energy were held constant. Conclusions/Applications: We conclude that using a modeling approach, the effect of diet on serum lipids in children is similar to that observed in adults. Total fat and saturated fat were positively associated with TC and HDL-C levels, saturated fat was positively associated with TC, and carbohydrate was inversely associated with both TC and HDL-C. In the statistical model, substitution of unsaturated fatty acids, monounsaturated fatty acids, or oleic acid for saturated fat, while holding total fat and energy constant, slightly lowered TC. In contrast, substitution of total fat for carbohydrate in the model increased TC and, thus, did not seem to be associated with an apparent health advantage except for HDL-C elevating effects. However, consumption of individual fats tends to be highly correlated, and we were unable to determine if these biological effects were operating independently.