Background: The clinical course of the neurally mediated syncope was known to be relatively benign but the effective treatment for these patients remained to be controversial. Objective: We retrospectively searched the efficacy of training, beta-blocker and midodrine in the neurally mediated syncope patients who were diagnosed through head-up tilt test. Methods: Two hundreds seventy two patients (M:F=107:165, mean age=36.4짹16.8) were enrolled and divided into 3 groups: group 1 (training group, n= 82, M:F=35:47, mean age=34.4짹17.6), group 2 (beta-blocker group, n=155, M:F=59:96, mean age=37.0짹17.0) and group 3 (midodrine group, n=35, M:F=13:22, mean age=38.3짹15.1). We compared the efficacy of three treatment modalities based on the recurrence rate of syncope and searched the adverse cardiac events. Results: There were no significant differences in baseline characteristics and types of neurally mediated syncope among these three groups. The recurrence rate of syncope was 3.7% in the group 1, 21.3% in the group 2 and 11.4% in the group 3 (p=0.0001) (see Table). There was no major cardiac event during the follow-up period. Conclusion: The recurrence rate of syncope is higher in beta-blocker group (21.3%) compared to training and midodrine group.