There has been no agreement among different authors on guidelines to specify the situations in which arthrodesis is justified in terms of results, risks and complications. The aim of this study was to identify preoperative predictors of outcome after decompressive lumbar surgery and instrumented posterolateral fusion. A prospective observational study design was performed on 203 consecutive patients. Potential preoperative predictors of outcome included sociodemographic factors as well as variables pertaining to the preoperative clinical situation, diagnosis, expectations and surgery. Separate multiple linear regression models were used to assess the association between selected predictors and outcome variables, defined as the improvement after 1 year on the visual analog scale (VAS) for back pain, VAS for leg pain, physical component scores (PCS) of SF-36 and Oswestry disability index (ODI). Follow-up was available for 184 patients (90.6%). Patients with higher educational level and optimistic preoperative expectations had a more favourable postoperative leg pain (VAS) and ODI. Smokers had less leg pain relief. Patients with better mental component score (emotional health) had greater ODI improvement. Less preoperative walking capacity predicted more leg pain relief. Patients with disc herniation had greater relief from back pain and more PCS and ODI improvement. More severe lumbar pain was predictive of less improvement on ODI and PCS. Age, sex, body mass index, analgesic use, surgeon, self-rated health, the number of decompressed levels and the length of fusion had no association with outcome. This study concludes that a higher educational level, optimistic expectations for improvement, the diagnosis of "disc herniation", less walking capacity and good emotional health may significantly improve clinical outcome. Smoking and more severe lumbar pain are predictors of worse results.