BACKGROUND: Vertebral metastases are the most frequent vertebral tumor. Transarterial embolization (TAE) devascularizes the tumor, resulting in tumor necrosis. Percutaneous vertebroplasty (PVP), a minimally invasive procedure, can effectively relieve tumor-related pain and improve spine stability. Unfortunately, the PVP technique is of limited use in controlling the progression of vertebral tumor, especially for paravertebral metastases. TAE combined with PVP may achieve a better control on vertebral metastases with paravertebral extension, but little information regarding the combination is available. OBJECTIVES: The present study is intended to assess the safety and effectiveness of the combination of TAE and PVP in patients suffering from vertebral metastases with paravertebral extension. STUDY DESIGN: Sequential TAE followed by PVP was used in 25 patients with symptomatic vertebral metastases. The safety and effectiveness of the sequential therapy were evaluated. SETTING: Three hospitals' clinical research centers. METHODS: This retrospective study was conducted with 25 consecutive patients (11 women and 14 men; mean age 59.3 years, range 38 - 80 years) with vertebral and paravertebral metastases from March 2009 to March 2014. The patients were treated with TAE, and 5 - 7 days later with the PVP procedure. The clinical outcomes were assessed by the control of pain using visual analog scale (VAS) scores, and computed tomography (CT) imaging. X2 or Fisher exact testing was performed for univariate analysis of variables. The VAS scores between groups were compared using ONE-WAY ANOVA, with a P-value of less than 0.05 considered statistically significant. RESULTS: All the TAE and PVP procedures were successfully done. Mean VAS scores decreased after TAE (from 8.64 ± 0.58 to 5.32 ± 1.46, P < 0.05) and further decreased after PVP (from 5.32 ± 1.46 to 2.36 ± 0.54, P < 0.05), and the decrease in VAS lasted until the third month (3.08 ± 1.52, P > 0.05) follow-up. However, VAS scores at the sixth month were statistically higher than those at the third month (4.8 ± 1.24 versus 3.08 ± 1.52, P < 0.05), VAS scores at the twelfth month were statistically higher than those at the sixth month (6.29 ± 1.07 versus 4.8 ± 1.24, P < 0.05). We found paravertebral cement leakage in 6 cases. No clinical or symptomatic complications were observed. In the follow-up, no patient showed further vertebral compression or spinal canal compromise. LIMITATIONS: This is a retrospective clinical study of a small number of patients. CONCLUSION: The sequential TAE followed by PVP is safe and effective in treating vertebral metastases with paravertebral extension. KEY WORDS: Spine, metastases, pain, embolization, vertebroplasty, interventional radiology, PVP, TAE.