In a prospective, randomised, placebo-controlled trial comparing the effect of (3-amino-1-hydroxypropylidene)-1,1-bisphosphonate (APD) (150 mg/day) plus calcium (1 g/day) with that of calcium alone on the bone mass of patients receiving long-term glucocorticoid therapy, the mean metacarpal cortical area in patients receiving APD increased by 1·2% between 0 and 6 months (p < 0·06) and then remained stable between 6 and 12 months. In contrast, this index progressively declined in the placebo group (p <0·05 at 12 months). The two groups differed significantly in the changes at both 6 and 12 months (p<0·01). Mean vertebral mineral density, as measured by quantitative computed tomography, increased by 19·6% over 12 months in the APD group (p<0·02) but showed a non-significant decline of 8·8% in controls. The differences between the changes were again significant (p<0·005). Biochemical indices and bone histomorphometry indicated a reduction in bone resorption and bone formation but there was no evidence of osteomalacia. APD may thus prevent bone loss in glucocorticoid-treated patients over a year.