Backgrounds: Infective endocarditis (IE) is associated with a high incidence of embolic events (EE), but surgical indications to prevent embolism remain controversial. We tried to compare clinical outcomes of early surgery to conventional management. Methods: From 1998 to 2006, we prospectively enrolled a total of 136 consecutive patients (89 men, age; 49짹17 years) with definite IE based on Duke criteria, and excluded patients with urgent indications of surgery based on current guidelines. The exclusion criteria were defined as patients with severe heart failure, aortic abscess, periannular complications, prosthetic valve IE, right-sided vegetations, small vegetations with diameter less than 5mm, or embolic stroke. The end point was defined as the composite of in-hospital death and clinical embolism confirmed by imaging studies. Results: Early surgery was performed on 67 patients (OP group) within 7 days of the diagnosis, and medical management was chosen on 69 patients (MED group). There were no significant differences between the two groups in terms of age, gender, euroSCORE and causative microorganisms, but the OP group had larger vegetations (16짹6 vs. 13짹5 mm, p=0.001) and a higher percentage of patients with severe valvular disease (90% vs. 51%, p<0.001). There were no EE, and 3 in-hospital deaths in OP group, and 13 EE and 2 in-hospital deaths in MED group (p=0.004). MED group (hazard ratio 9.09, p=0.002) and vegetation size (hazard ratio 1.10, p=0.04) were independently associated with in-hospital events on multivariate analysis. In the MED group, 41 (59%) patients underwent late valve surgery after effective antimicrobial therapy. During median follow-up of 36 months, there were 1 death, 2 EE and 1 recurrence of IE in the MED group, and 1 EE in the OP group. The 3-year event-free survival rate was significantly higher in the OP group (93짹3%) than in the MED group (71짹5%, p=0.001). For the 42 propensity score-matched pairs, the OP group had the lower in-hospital event rate (hazard ratio 0.10, p=0.037), and the higher event free survival rate at 3 years (93짹4% vs 79짹6%, p=0.024). Conclusion: Early surgery is associated with more improved clinical outcomes than a conventional treatment strategy via a significant decrease in embolic events.