Whether vitamin D receptor (VDR) gene polymorphisms are associated with osteoporosis is highly controversial. The relationship between VDR gene polymorphisms and bone mineral density (BMD) might, however, be modified by age-related and/or environmental factors. We studied the potential association between BMD and VDR genotypes in females from prepuberty to premenopause and prospectively investigated the interaction of VDR genotypes with dietary calcium and BMD changes during childhood. Bsm I VDR gene polymorphisms and BMD at the lumbar spine (LS) and femur (neck [FN] and midshaft [FS]) were assessed in 369 healthy Caucasian females, aged 7-56 years (143 prepubertal girls, 54 peri- and postpubertal adolescents, and 172 premenopausal adults). Femoral trochanter (FT) and distal radius BMD (metaphysis and diaphysis) were also measured in 101 of the prepubertal girls who participated in a 1-year, double-blind, randomized study of calcium supplementation (850 mg/day) versus placebo on bone mineral mass accrual. Among all females, 150 (40.7%) had bb, 167 (45.3%) Bb, and 52 (14%) BB VDR genotypes. In prepubertal and adolescent girls altogether, LS BMD (Z scores) was associated with VDR genotypes and was significantly lower in BB than in Bb or bb subjects. Trends for a similar difference were also detected at the FN level as well as on the mean BMD (Z scores) of the three sites measured (LS, FN, and FS). By contrast, no BMD differences were detectable among VDR genotypes in the adults. In 101 prospectively studied prepubertal girls, calcium supplementation significantly increased BMD at most skeletal sites, except LS. After segregation for VDR genotypes (40 bb, 47 Bb, and 14 BB), a significant calcium effect was present in Bb but not bb girls, whereas in BB girls there was a positive but nonsignificant trend for a calcium effect. Moreover, dietary calcium intake was significantly correlated with BMD changes at various independent bone sites in Bb girls but not in bb girls. In contrast, BMD gain in bb girls appeared to be higher than among the other genotypes when the dietary calcium intake was low, i.e., in the absence of calcium supplements. BMD was significantly associated with VDR gene polymorphisms only before puberty, BB girls having significantly lower BMD (Z scores) than the other genotypes. By increasing dietary calcium intake, BMD accrual was increased in Bb and possibly BB prepubertal girls, whereas bb subjects had the highest spontaneous BMD accrual and remained unaffected by calcium supplements. Taking into account complex interactions between VDR gene polymorphisms and environmental factors, including calcium intake, may thus help to understand the discordant relationships between BMD and VDR gene polymorphisms.