Very low-birth weight (VLBW) infants fed high-calcium and high-phosphorus (10.74 and 6.93 mmol/MJ; 180 and 90 mg/100 kcal, respectively) infant formulas were randomized to one of three levels of vitamin D intake to approximate 200, 400, and 800 IU/day. Sixty-two infants completed the study (24 to 29 days), with actual mean daily vitamin D intakes of 161, 361, and 766 IU, respectively. Outcomes were not different by group: gains in body weight, length and head circumference, serum calcium, magnesium, phosphorus, alkaline phosphatase, osteocalcin, 25-hydroxyvitamin D, 1,25-dihydroxyvitamin D, and urine calcium:creatinine and magnesium:creatinine ratios. There were no radiographic fractures and/or rickets. A subset of 19 infants was followed between 173 and 380 days to determine descriptively if there was any delayed effect of earlier manipulation of vitamin D intake. They were fed standard infant formulas with a vitamin D content of 400 to 420 IU/L. No significant differences were present among the three groups, and data were combined. Serum 25-hydroxyvitamin D increased (p < 0.05), osteocalcin decreased (p < 0.05), and 1,25-dihydroxyvitamin D decreased (p = 0.06) at follow-up. Thus, for VLBZW infants fed high-calcium and high-phosphorus milk, an average daily vitamin D intake as low as 160 IU maintains normal and stable vitamin D status and normal physical growth, biochemical and hormonal indexes of bone mineral metabolism, and skeletal radiographs versus randomized infants receiving about 400 or 800 IU of vitamin D per day. On follow-up, vitamin D status remained normal for > or = 6 months while infants received < 400 IU of vitamin D per day.