This 2-year study was carried out in 36 healthy women of mean age 53.9 +/- 3.8 (SD) years and 3.4 +/- 2.3 years postmenopausal. Bone mineral density (BMD) in the spine, measured by single-energy quantitative computed tomographic scanning, gave a mean initial value of 110 +/- 26 mg/ml. The women were divided randomly into group 1 (n = 11), calcium 600 mg/day; group 2 (n = 15), calcium plus etidronate sodium 400 mg/day for 14 days every 3 months; and group 3 (n = 10), calcium plus etidronate plus phosphate, the 14-day etidronate course being preceded by phosphate 1 g twice daily for 3 days. During the first year of the study BMD decreased by 6.0 +/- 5.8% (p < 0.005) in group 1 subjects and increased by 4.5 +/- 7.8% (p < 0.005) in the combined etidronate-treated groups (difference between control and treated p < 0.001). Inclusion of phosphate in the regimen did not affect the response to etidronate. In the second year there was no significant mean change in BMD in any of the three groups. However, whilst there was little change in BMD values for most of the group 1 subjects, there was considerable variation in individual response within the etidronate-treated groups, with some subjects gaining and some losing bone. The change in BMD during the second year in the subjects as a whole was highly correlated with the change in plasma calcium after 3 months of treatment (r = 0.60, p < 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)