The prevention and treatment of glucocorticoid-induced osteoporosis is a major concern for rheumatologists since inflammatory joint disease is among the most common reasons for long-term glucocorticoid therapy. We used a randomized placebo-controlled design to evaluate the efficacy of one-year cyclical etidronate therapy in preventing bone loss in 83 glucocorticoid-treated patients with rheumatoid arthritis, polymyalgia rheumatica, or giant cell arteritis. Glucocorticoid treatment duration was shorter than three months, and the starting dose was greater than 7.5 mg of prednisone-equivalent per day. Etidronate was given according to the standard cyclical schedule, i.e. 400 mg/d for periods of 14 days separated by 76-day intervals during which patients took 500 mg of supplemental calcium per day. The primary evaluation criterion was the change in lumbar spine bone mineral density after one year of etidronate therapy. Bone mineral density decreased by 1.94 +/- 0.61% in the placebo group and increased by 0.86 +/- 0.6% in the etidronate group, yielding a between-group difference of 2.8 +/- 0.86% (P = 0.002). The difference was largest in postmenopausal women (3.38 +/- 1.11%; P = 0.004). At the femoral neck, there was a smaller bone mineral density decrease in the etidronate than in the placebo group, but the difference (1.11 +/- 1.13%) was not statistically significant. The most common side effects were gastrointestinal symptoms and showed no difference between the two groups. Four fractures (including one vertebral fracture) occurred in the placebo group versus two (including one vertebral) in the etidronate group. Etidronate prevents glucocorticoid-induced lumbar spine bone loss in patients with rheumatoid arthritis, polymyalgia rheumatica, or giant cell arteritis.