BACKGROUND The objective of this study was to analyse the impact of acute surgery for native aortic valve endocarditis and its influence on the long-term prognosis after surgery. METHODS A total of 161 patients underwent aortic valve replacement for native active aortic valve endocarditis (NAAVE) during a 29-year period, from 1967 to 1995 (age range: 10 to 72 years; mean 48 +/- 12). The main indication for surgery was progressive congestive heart failure (76%). Other indications were untreatable sepsis (27%), peripheral or central emboli (12%) and, from 1978, echocardiographic evidence of friable, pedunculated vegetations (3%). Streptococcal and staphylococcal infections predominated. Concomitant procedures were performed in 27% of the patients, including mitral and tricuspid valve surgery and coronary bypass procedures. RESULTS Operative mortality was 8% in the majority of cases caused by heart failure or multi-organ failure. Multivariate logistic regression analysis identified NYHA class IV to be an independent predictor for postoperative death. Long-term survival for discharged patients was 75% at 10 years and 58% at 15 years, with a mortality rate of 3.6%/patient/year. Cox regression analysis identified the year of operation, trivalvular endocarditis and staphylococcal infection as independent predictors of survival. At 10 and 15 years after aortic valve replacement, 91% and 84% of the patients, respectively, were free of recurrent endocarditis. The presence of an abscess cavity at first operation was found to be predictive of recurrent endocarditis. CONCLUSIONS Valve replacement for NAAVE offers a good chance for a cure and satisfactory long-term survival. Improvements in pre- and per-operative management of the very ill patient, and the use of allograft valves are likely to further improve long-term results. Finally, the presence of staphylococcal endocarditis requires long-term postoperative antibiotic therapy.