The effects of three isocaloric intravenous nutritional regimens were studied in seven infants and children, ages 2 months to 9 yr, with congenital gastrointestinal anomalies (four patients) or with prior history of malignant disease admitted in remission for bone marrow transplantation (three patients). Energy metabolism, as measured by the basal metabolic rate (BMR), and substrate utilization, as measured by the respiratory quotient (RQ), were studied to determine the effect of different levels of carbohydrate and fat on nitrogen retention in each patient. Solution A provided 8% of energy as amino acids, 87% as carbohydrate, and 5% as fat. Solution B provided 8% of energy as amino acids, 60% as carbohydrate, and 32% as fat. Solution C provided 8% of energy as amino acids, 34% as carbohydrate, and 58% as fat. Administration of solution A (high carbohydrate, low fat) was associated with moderately increased mean (+/- SD) BMR and RQ and with low nitrogen retention (19.1 +/- 12.7%, 1.06 +/- 0.14, and 98 +/- 28 mg N/kg/day). Both the BMR and the RQ decreased when less carbohydrate and more lipid was given: BMR 4.3 +/- 11.6% (p less than 0.005), RQ 0.92 +/- 0.09 (p less than 0.001) for solution B; BMR 3.94 +/- 10.6% (p less than 0.005), RQ 0.86 +/- 0.09 (p less than 0.001) for solution C. Among the solutions tested, optimal nitrogen retention [163 +/- 60 mg N/kg/day (p less than 0.01)] was noted with solution B. Our data support the conclusion that a physiologic balance of fat and carbohydrate results in optimal nitrogen retention.