Our previous studies raised two hypotheses: first that suboptimal early nutrition and second that human milk have enhancing effects on long-term bone mineralization. To test these hypotheses experimentally, we measured whole body and regional bone mineral content (BMC) and bone mineral density (BMD), using dual-energy X-ray absorptiometry and single-photon absorptiometry, and bone turnover at 8-12 years in 244 preterm children (128 boys) who participated in a prospective randomized study of diet during the neonatal period. Dietary randomizations studied were: banked human milk (BBM, n = 87) versus preterm formula (PTF, n = 96) as the sole diet or as a supplement to mother's expressed breast milk (EBM); PTF (n = 25) versus term formula (TF, n = 36) as sole diet. Ninety-five term children of the same age were also studied. First, preterm children were shorter and lighter than term children (height SD scores -0.49 (1.1) vs. +0.22 (0.9), weight SD scores -0.41 (1.2) vs. +0.38 (1.0)) and had significantly lower whole-body BMC than their peers; decrements were also evident at some regional sites. These differences disappeared after adjusting for bone area, body size, and pubertal status. Second, children previously randomized to BBM versus PTF or TF versus PTF showed no significant differences in anthropometry, BMC, BMD, or osteocalcin (OC). Third, there was no independent effect of the proportion of EBM on BMC, BMD, or OC and no interaction between randomized diet and the amount of EBM received. Fourth, plasma OC was significantly higher in preterm children than in term children (12.4 vs. 11.0 ng/ml, p < 0.005) and in preterm children who had received a low-nutrient (BBM/TF) as opposed to a high-nutrient diet (PTF) during the neonatal period (12. 9 vs. 11.9 ng/ml, p = 0.03). In conclusion, preterm children are shorter, lighter, and have lower bone mass than their peers at age 8-12 years. The lower BMC is, however, appropriate for the bone and body size achieved. Despite large differences in early mineral intake, early diet does not affect bone mass in preterm children, and fresh human milk has no specific effect. However, poor nutrition during the neonatal period may result in higher bone formation rates during childhood.