We tested the spine antifracture and bone sparing efficacy of 1.2 g/day of oral calcium as carbonate in two groups of elderly women, one with prevalent fractures (PF, n = 94) on entry and the other without (NPF, n = 103). It was a prospective randomized, double-blind, placebo-controlled trial in mostly rural communities in women over age 60 who were living independently and were consuming < 1 g/day of calcium. We obtained annual lateral spine radiographs and semiannual forearm bone density over 4.3 +/- 1.1 years and determined vertebral fractures by radiographic morphometry augmented by physician assessment. In the PF group, 15 of 53 subjects on calcium had incident fractures, compared with 21 of 41 on placebo (p = 0.023, chi2). Calcium did not reduce the rate of incident fractures in the NPF group. Those with a prevalent fracture on entry and not treated with calcium were 2.8 times more likely to experience an incident fracture than all others. Change in the forearm bone mass on placebo in the PF group was -1.24 +/- 2.41%/year compared with +0.31 +/- 1.80%/year on calcium (p < 0.001). In the NPF group, the difference was less: -0.39 +/- 2.08%/year versus 0.00 +/- 1.64%/year (p = 0.2). We conclude that in elderly postmenopausal women with spine fractures and selfselected calcium intakes of < 1 g/day, a calcium supplement of 1.2 g/day reduces the incidence of spine fractures and halts measurable bone loss.