We report the outcome of 71 consecutive posterolateral lumbar fusions without spinal instrumentation. The indication for the operation was spondylolysis-olisthesis, degenerative disc disease/facet joint arthrosis, or pain after prior laminectomy. Concerning pain relief, 29/43 patients with spondylolysis-olisthesis were classified as good. The corresponding figures in the group with degenerative disc disease and/or facet joint arthrosis were 8/16 patients and in the group with pain post-laminectomy, 6/12 patients. No surgical complications were noted. In the total material 54 patients had a solid fusion, as defined by radiographic osseous trabecular bridging at all intended levels. One-level fusions tended to heal solidly in a higher frequency than two-level fusions. For the spondylolysis-olisthesis group, healed fusion correlated with a good clinical result. Such a correlation could not be verified for the other diagnostic groups. We conclude that non-instrumented posterolateral lumbar fusion is a valid method for treating low-grade spondylolysis-olisthesis, especially when the aim is to fuse a single level. Improved patient selection methods are required in fusion for degenerative disc disease and pain after laminectomy.