Background Aortic valve surgery for endocarditis remains a high-risk procedure. The objective of this study was to analyze the interaction between the various subsets of endocarditis (native, prosthetic, healed, and active), timing of surgery, and their influence on early and late outcomes.Methods and Results During a 20-year period starting January 1972, 200 patients underwent aortic valve replacement for infective endocarditis (age range, 13 to 88 years; median, 53 years). There were 51 (26%) females, and 109 (55%) were in New York Heart Association functional class IV before surgery. Native valve endocarditis (NVE) and prosthetic valve endocarditis (PVE) were present in 132 (66%) and 68 (34%) patients, respectively. Surgery was required in 120 (60%) during the active phase (AE) and 80 (40%) during the healed phase (HE) of endocarditis. The main indication for surgery in the healed group was progressive congestive heart failure. The indications for the active group were congestive heart failure (68%), continuing active sepsis (70%), echocardiographic vegetation (28%), peripheral emboli (30%), and arrhythmias (13%). Streptococcal infections predominated in NVE, staphylococcal in PVE and AE; culture-negative endocarditis predominated in the healed group. Isolated aortic valve surgery was performed in 68% of the patients, and concomitant procedures (32%) included mitral valve and coronary bypass procedures. The overall operative mortality (OM) was 12.5%. The OM was 7.5% and 22% for NVE and PVE, respectively (P=.004), and 7% for HE versus 15% for AE (P=.06). The OM for early PVE was 33% versus 18% for late PVE (P<.05). Multivariate logistic regression analysis identified PVE and New York Heart Association functional class IV to be independent predictors for early death. Recurrent endocarditis occurred 26 times in 24 patients (11 early, 13 late), with three operative deaths in the early group, all due to residual staphylococcal infections. Freedom from recurrent endocarditis was significantly different between HE (96+-3% and 86+-6% at 5 and 10 years, respectively) and AE (89+-3% and 83+-4%, respectively) (P=.02). Long-term survival for discharged patients was 81+-3% and 63+-5% at 5 and 10 years, respectively, with no significant difference between NVE, PVE, AE, and HE.Conclusions These data suggest that for active endocarditis, surgery should be delayed to achieve a healed status provided there is no pressing need for immediate surgery.Patients with staphylococcal endocarditis, particularly on a prosthesis, should be operated on sooner and should be covered with antibiotics for an extended period to prevent recurrent PVE. This study stresses the need for aggressive antibiotic prophylaxis, particularly in the presence of a prosthesis. (Circulation. 1994;90(part 2):II-175-II-182.)