To evaluate the effectiveness of intravenous clodronate in ameliorating refractory bone pain in patients with metastatic bone disease, 60 patients with established osseous metastases and persistent bone pain were randomized to receive either clodronate (600 mg or 1500 mg in 500 mL of normal saline) or 500 mL of saline as placebo. After 2 weeks, the patients were crossed over to receive the alternate treatment. After another 2 weeks, each patient and investigator made a blinded choice. Daily visual analogue scales (VAS) and analgesic diaries were recorded throughout the study period. Forty-six patients were evaluable (77%). A treatment x period interaction was identified in the VAS and daily morphine equivalent dose (DMED) scores. First period analysis of the VAS scores for general pain, pain at rest, and pain upon movement demonstrated an average reduction of 13, 14, and 24 mm, respectively, from baseline, but were not significantly different from changes following placebo. The average change in DMED was -6.4 (SE = 2.9) following clodronate and was +24.6 (SE = 14.9) following placebo (p = 0.03). In the blinded choice of which agent resulted in improvement in pain, 26 (57%) patients chose clodronate, 12 (26%) chose placebo, and eight (17%) had no preference (p = 0.0021). For the investigators who also made a blinded selection, clodronate was chosen in 30 (65%) patients, placebo in ten (22%) patients, and no difference was apparent in six (13%) (p < 0.0001). Intravenous clodronate appeared to have analgesic effect in patients with refractory bone pain due to metastatic bone disease. The optimal dose and duration of effect require further evaluation, particularly in patients with stable disease and persistent bone pain.