A genetic marker for the 1,25-dihydroxyvitamin D receptor (VDR) is reported to account for much of the heritable component of bone density. It is not known whether VDR genotype influences bone accretion or loss, or how it is related to calcium metabolism. The VDR genotype was determined in 229 healthy postmenopausal women who previously participated in a calcium trial. VDR alleles were designated according to presence (b) or absence (B) of the BsmI restriction enzyme cutting site. There were 83 bb, 102 Bb, and 44 BB individuals. Two-thirds of the women took 500 mg of calcium supplement (mean calcium intake = 892 mg/day) and one-third a placebo (mean = 376 mg/day). Bone mineral density (BMD) at the femoral neck, spine, and radius were measured by dual- and single-photon absorptiometry at baseline and after 1 and 2 years. Among women more than 10 years postmenopausal, those with the BB genotype had the lowest femoral neck BMD. Rates of bone loss over 2 years were greater in the BB group at all sites (e.g., at the femoral neck, bb, 0.45 ± 0.43; Bb, -0.01 ± 0.40; BB, -0.99 ± 0.50%/year; BB vs. bb, p = 0.01), and this trend was found both in women <10 years since menopause (e.g., at the radius, bb, 0.43 ± 0.47; Bb, -0.37 ± 0.42; BB, -1.20 ± 0.59% per year; BB vs. bb, p = 0.02) and those ≥10 years (radius, bb, -0.71 ± 0.41; Bb, 0.08 ± 0.39; BB, -1.41 ± 0.49% per year; BB vs. Bb, p < 0.01). At the femoral neck, bone loss appeared to be modified by calcium intake (e.g., in the BB genotype: +0.03 ± 0.61 in supplemented vs. -2.01 ± 0.75%/year in placebo, in bb: 0.57 ± 0.58 vs. 0.32 ± 0.47%/year; interaction term p = 0.09), and this trend was also present in both early and late menopause. Rates of change at the radius and spine in BB were not significantly influenced by calcium at the intake levels of this study group. These results indicate that postmenopausal bone loss is influenced by the VDR genotype and suggest the adverse effect of the susceptible allele at the hip may be reduced by raising calcium intake.