The purpose of this study was to see whether there were any correlations between the clinical findings and outcomes after the epidural steroid injection(ESI), and to determine which of the clinical findings might be of importance as favorable signs to predict a good result with the ESI. The present study involves 23 elderly patients(18 females and 5 males) with spinal stenosis. The participants were divided into several subgroups according to the clinical features, such as presenting symptoms, age, duration of symptoms and physical findings. Subjects received the epidural injection of 80 mg Depo-Medrol with 6 ml of 1.5% lidocaine solution. The subjects completed a questionnaire before the ESI, and at three weeks and two months after the ESI respectively. The questionnaire included three scales on symptom severity, physical function, and satisfaction. The visual analogue pain scale was utilized as well. At three weeks, satisfaction outcomes were encountered in 82.6 percent of the sudjects. Differences in the satisfaction between the subgroups according to the duration of symptoms and abnormal physical findings were statistically significant(P<0.05). The total scores of symptom severity and physical function had improved from the baseline values(P<0.0001). However, the score changes in each subgroup, for the duration of symptoms differ statistically same (P<0.05) as assessed by the visual analogue pain scale. The results at two months were similar to those of at three weeks. Approximately 78.3% of the patients reported the satisfactory results. Although the follow up period was short, the results demonstrated that the epidural steroid injection afforded much improvement in clinical symptom and physical function in elderly patients with the lumbar spinal stenosis. The only significant difference was the satisfaction outcomes in the groups with short duration of symptoms. The determination of persisting outcomes from the ESI would require a long-term follow-up study.