Thirty six preterm infants (20 boys) of 25 to 32 weeks' gestation were observed from birth to around 40 weeks' postconception. When oral feeding became possible, nine received mother's own breast milk (group B), 15 formula feed (group F), and 12 formula feed supplemented with calcium (5 ml 10% calcium gluconate/100 ml feed) and phosphorus (0.5 ml 17% potassium phosphate similarly) (group S). All received a daily supplement of 400 IU vitamin D. Intakes of calcium, phosphorus, vitamin D, energy, and fluid volume were recorded. When oral feeding started, and near 40 weeks' postconception, bone mineral content of the forearm was measured by photon absorptiometry; weight and crown-heel length were also measured. After logarithmic transformation of the measurements, there were no significant intergroup differences between the mean rate constants for weight or crown-heel length describing growth during the observation period. The mean rate constant for mineral accretion (M) was significantly higher in group S than in both the others. Pooling all data, M was significantly correlated with calcium intake but not with any other variable. Mineral supplementation of feed can reduce but not cure osteopenia of prematurity.