We hypothesized that retention of parenterally delivered calcium (Ca) and phosphorus (P) is affected by the ratio of the delivered minerals and that a 1.7:1 ratio would be optimal since this is the ratio of retention of these minerals by the fetus. Forty-one very low birth weight (VLBW) infants were randomly assigned to one of three total parenteral nutrition (TPN) solutions that were different only in their Ca:P ratios: 2:1 (76 mg/kg/day Ca and 38 mg/kg/day of P), and 1.3:1 (58 mg/kg/day Ca and 45 mg/kg/day P), and 1.3:1 (58 mg/kg/day of Ca and 45 mg/kg/day of P). Serum levels of calcium, phosphorus, and alkaline phosphatase, retentions of calcium and phosphorus and urinary cyclic AMP levels were measured after 48 h on the assigned Ca to P ratio. Calcium retentions were higher with the 2:1 and 1.7:1 ratios and phosphorus retentions were higher with the 1.3:1 and 1.7:1 ratios. The 1.7:1 ratio allowed for the highest absolute retention of both minerals and was the closest to published in utero accretion of calcium and phosphorus. The serum and urine studies demonstrated no abnormalities on any of the three ratios. Cyclic AMPs were not different among groups and were not elevated compared to previous reports suggesting that none resulted in parathyroid hormone (PTH) stimulation. We conclude that the 1.7:1 ratio is better than higher or lower ratios for delivery of calcium and phosphorus in TPN solutions at the quantities studied.