Lumbar spine fusion is a common procedure associated with a high cost burden and risk of serious complications. We aimed to summarise systematic reviews on the effectiveness of lumbar spine fusion for most diagnoses. We found no highâ€quality systematic reviews and the risk of bias of the randomised controlled trials in the reviews was generally high. The available evidence does not support a benefit from spine fusion compared to nonâ€operative alternatives for back pain associated with degeneration. The available evidence does not support a clinical benefit from spine fusion compared to nonâ€operative treatment or stabilisation without fusion for thoracolumbar burst fractures. Benefits of spine fusion compared to nonâ€operative treatment for isthmic spondylolisthesis are unclear (one trial at high risk of bias). Surgical intervention for metastatic carcinoma of the spine associated with spinal cord compromise improves mobility and neurological outcome (based on a single trial). Better evidence is required to determine more accurately the effectiveness of spine fusion surgery for all indications. Patients contemplating spinal fusion should be fully informed about the evidence base for their particular problem, including the relative potential benefits and harms of fusion compared with nonâ€operative treatments.