The aim of this study was to assess the efficacy of percutaneous cementoplasty (PC) with polymethylmethacrylate (PMMA) in painful extravertebral lytic bone metastases not responding to conventional therapy. Fifty patients (25 females), mean age 64.7 +/- 11.2 years, underwent PC after giving informed consent. Procedures were performed under fluoroscopy (1/50) or combined fluoroscopy-CT (49/50) guidance in local anesthesia or under deep sedation in 7 patients with large metastases who underwent radiofrequency thermoablation (RFA) in the same session. Seventy lesions were treated (1-6 per patient; average, 1.4 +/- 0.9), arranging in size from 1 to 10 cm (average, 3.6 +/- 2.1 cm). Mean volume of PMMA per lesion was 5.9 +/- 3.2 ml (range, 1.5-15.0 ml). Pain was prospectively evaluated on an 11-point visual analog scale (VAS) before and after the procedure (follow-up, 15 to 36 months). Mean VAS score dropped from 9.1 +/- 1.2 (range: 6-10) to 2.1 +/- 2.5 (range: 0-9). Mean VAS difference was 7.0 +/- 2.3 (range, 1-10; p \ 0.0001, Wilcoxon signed rank test). Forty-seven of the 50 patients (94%) suspended narcotic drugs, in 22 (44%) pain was controlled with a nonsteroidal anti-inflammatory drug, in 25 (50%) analgesic therapy was suspended, and 13 of 50 (26%) had complete pain regression. In 3 of the 50 patients (6%) pain was not improved. No statistical difference between osteoplasty and osteoplasty plus RFA was found (p = 0.8338, Mann-Whitney test). No complications arose during the procedure. Two patients with metastases in the femoral diaphysis reported a fracture 1 month after treatment. PC is effective to obtain pain regression in painful bone metastases not responding to conventional analgesic therapy; bone consolidation cannot be obtained in the diaphysis of long weight-bearing bones.