We have performed a 2-year prospective double-masked study to determine whether the bisphosphonate pamidronate can prevent bone loss in postmenopausal women and its optimal dosage regimen. One hundred and twenty-one such women (mean +/- SD age 57.6 +/- 3.4 years; mean +/- SD time since menopause 7.5 +/- 3.5 years) were randomized to receive either oral pamidronate (300 mg/day) for 4 weeks every 4 months (group A), oral pamidronate (150 mg/day) for 4 weeks every 2 months (group B) or identical placebo capsules (group C). Bone mineral density (BMD) measurements at the lumbar spine and proximal femur were performed at baseline and at 6-month intervals for 2 years using dual-energy X-ray absorptiometry. BMD at the lumbar spine (L2-4) increased significantly in groups A and B after 2 years of treatment (mean +/- SD 2.8 +/- 2.1% and 3.0 +/- 2.9% respectively, both p < 0.001) but decreased in the placebo group (-1.6 +/- 3.1%, p < 0.01). Identical results were seen for BMD at the femoral neck, which increased significantly in groups A and B after 2 years of treatment (1.2 +/- 2.3% and 1.3 +/- 2.9% respectively, both p < 0.05) but decreased in the placebo group (-1.9 +/- 3.9%, p < 0.05). There were significant differences over 2 years between the groups at all anatomical sites (lumbar spine, femoral neck and trochanteric region, all p < 0.001; Ward's triangle, p < 0.01). However, there were no significant differences between groups A and B, suggesting that the two treatment regimens were equally effective in conserving BMD. There were, however, marked differences in tolerability between the two treatment regimens: 13 women (34%) in group A withdrew from the study because of side-effects, but only 5 women (12%) in group B, which was comparable with placebo. These data demonstrate that intermittent oral pamidronate will prevent bone loss from the lumbar spine and proximal femur of postmenopausal women, and that the more frequent but lower dose regimen is well tolerated.