To determine the biological criteria for neonatal vitamin D deficiency, serum 25-hydroxyvitamin D (calcidiol), parathyroid hormone (PTH), calcium, phosphates, and alkaline phosphatase (ALP) activity were measured during the winter-spring period in 80 healthy neonates and their mothers 3-6 d after delivery. A longitudinal 3-mo survey of the serum biology of 52 of these neonates consuming formula was also performed to test the influence of their neonatal vitamin D status on the effects of two oral ergocalciferol supplements (500 and 1000 IU or 12.5 and 25 micrograms/d). At birth, 63.7% of the infants had calcidiol concentrations < or = 30 nmol/L. Most of them had no other biological sign evocative of vitamin D deficiency, but 14 neonates had low calcidiol concentrations and serum PTH concentrations > 60 ng/L, the upper limit of the adult normal range. They also had a significantly lower mean serum calcium concentration than did neonates with calcidiol concentrations > 30 nmol/L. On the basis of the association of low calcidiol concentrations (< or = 30 nmol/L) and high PTH concentrations (> 60 ng/L) as criteria for vitamin D deficiency, 24% of the neonates born to unsupplemented mothers were found to be vitamin D-deficient. Neonatal vitamin D status influenced the response of the infants to vitamin D supplements. Neonates with no sign of vitamin D deficiency showed similar changes in their serum calcidiol, calcium, phosphate, and PTH concentrations and ALP activity and no toxic effect (hypercalcemia or highly elevated calcidiol concentration) was observed whatever their vitamin D intake. In contrast, neonates with subclinical vitamin D deficiency had normalized serum PTH within 1 mo only when they were given 1000 IU ergocalciferol (25 micrograms)/d in addition to their formula.