The functional outcome of patients with lumbar herniated nucleus pulposus without significant stenosis was analyzed in a retrospective cohort study. Inclusion criteria were as follows: 1) a chief complaint of leg pain, primarily; 2) a positive straight leg raising (SLR) at less than 60 degrees reproducing the leg pain; 3) a computed tomography (CT) scan demonstrating a herniated nucleus pulposus without significant stenosis by a radiologist's reading, which was also confirmed by the authors; 4) a positive electromyogram (EMG) demonstrating evidence of radiculopathy; and 5) response to a follow-up questionnaire. All patients had undergone an aggressive physical rehabilitation program consisting of back school and stabilization exercise training. Of a total of 347 consecutively identified patients, 64 patients with an average follow-up time of 31.1 months met the inclusion criteria and constituted the study population. They were sent questionnaires that inquired about activity level, pain level, work status, and further medical care. The patients with neurologic loss, extruded discs, and those seeking a second opinion regarding surgery were identified and subgrouped. Results for the total group included 90% good or excellent outcome with a 92% return to work rate. For the subgroups with extruded discs and second opinions, 87% and 83% had good or excellent outcomes, respectively, all (100%) of whom returned to work. Sick leave time for these subgroups was 2.9 months (+/- 1.4 months) and 3.4 months (+/- 1.7 months), respectively. These results compared favorably with previously published surgical studies. Four of six patients who required surgery were found to have stenosis at operation.(ABSTRACT TRUNCATED AT 250 WORDS)