Aim: To establish the efficacy and complications associated with vertebroplasty in spinal metastases and myeloma. Materials and methods: A literature search was performed from inception to April 2010. Thirty relevant studies were identified. Only one was a randomized, controlled trial and seven were prospective studies. Nine hundred and eighty-seven patients aged between 45 and 72 years were included in this systematic review. Results: Most studies report performing the procedure under local anaesthetic and continuous fluoroscopic screening, and only two centres reported treating more than four vertebrae per session. Five deaths were attributable to vertebroplasty, with a further 19 patients suffering a serious complication related to the procedure. There is some evidence to suggest that the complication rate may be related to the higher cement volume used, although the data are not robust enough for meta-analysis. Pain reduction ranged between 47-87%, similar to the results for osteoporosis. There was no correlation between pain reduction and cement volume. Conclusion: This systematic review reveals the paucity of good-quality, robust data available on the subject of percutaneous vertebroplasty in malignancy. It also highlights the apparent high risk of serious complication (2%). Further research into the subject is required in this group of patients.