The clinical course of 19 untreated patients with spinal stenosis (mean age, 60 years) was compared with that of 44 patients treated surgically (mean age, 65 years). The time of follow-up was 31 and 53 months, respectively. About 80% of the patients had neurogenic intermittent claudication. In the follow-up, one third of the treated and one half of the untreated patients still had neurogenic claudication. By visual analogue-scale estimation, 60% of those treated surgically and 33% of the untreated patients felt better. Fifty-eight percent of the untreated patients were unchanged. Neurophysiologic changes showed progression in almost all cases; it was more pronounced in the treated patients. No proof of severe deterioration was found in the untreated patients, and observation for 2-3 years seems to be a good alternative to surgery.