We randomised a total of 94 patients with long-standing moderate lumbar spinal stenosis (LSS) into a surgical group and a non-operative group, with 50 and 44 patients, respectively. The operative treatment comprised undercutting laminectomy of stenotic segments, augmented with transpedicular-instrumented fusion in suspected lumbar instability. The primary outcome was the Oswestry disability index (ODI), and the other main outcomes included assessments of leg and back pain and self-reported walking ability, all based on questionnaire data from 85 patients at the 6-year follow-up. At the 6-year follow-up, the mean difference in ODI in favour of surgery was 9.5 (95% confidence interval 0.9-18.1, P-value for global difference 0.006), whereas the intensity of leg or back pain did not differ between the two treatment groups any longer. Walking ability did not differ between the treatment groups at any time. Decompressive surgery of LSS provided modest but consistent improvement in functional ability, surpassing that obtained after non-operative measures.