The present study focuses on the long-term prognosis of radiographically verified stenosis of the lower lumbar spine. The purpose here was to describe the outcome 12 years after radiographic diagnosis of spinal stenosis and to identify factors predicting disability after operative or conservative treatment. Data were compiled on 75 patients (43 men and 32 women) with changes in functional myelography diagnostic for spinal stenosis. Their mean age at the interview 12 years later was 61 years. The sagittal diameter of the dural sac was measured from baseline myelographs at all intervertebral levels and was corrected for magnification. In the interview, subjective outcome assessment was obtained with a structured questionnaire, and the low-back disorder was scored using the Oswestry disability index. The sagittal diameter of the dural sac was severely stenotic (<7.0 mm) in 32 patients (26 operated), and moderately stenotic (7.0-10.5 mm) in 43 patients (31 operated). The severity of the stenosis significantly predicted disability, even when the effects of age, sex, therapy regimen, and body mass index were adjusted for. For moderate and severe stenosis, the adjusted mean Oswestry indices were 28.4 and 39.1, respectively (p = 0.01). Therapy as such (operative versus nonoperative) did not significantly correlate with later disability. The radiographic severity of lumbar spinal stenosis predicts disability independently of therapy regimen. Randomized clinical trials are needed to establish the indications for surgical and conservative treatment. Radiographic severity of the stenosis should be considered as an effect-modifying or confounding factor in clinical trials and other studies focusing on the outcome of lumbar spinal stenosis.