The outcome in 132 patients with infective endocarditis diagnosed in accordance with the Duke criteria at a tertiary hospital in Denmark in the period 1998-2000 is reported. The total in-hospital mortality was 15%. Indications are that in-hospital mortality over the last decade has been reduced by around a quarter. Mortality after 3 months was 17% (CI 29%), after 3 years 32% (CI 16-47%) and after 5 years 39% (CI 22-55%). This 5-years mortality was 5 times that of an age and gender matched background population. After follow-up for 5-8 y, mortality was highest for prosthetic valve endocarditis (63% vs. 39%, p=0.05). Heart surgery was performed in 51% of the cases. Patients who underwent surgery had a lower mortality at follow-up (36% vs. 52%, p=0.04). The 5-year mortality was 30% (CI 9-52%) for patients treated with surgery and 48% (CI 23-72%) for patients treated without surgery. In multivariable analysis surgery was not an independent predictor for lower long-term mortality. Surgery was however an independent predictor for lower intermediate-term mortality. It is concluded that surgery may be associated with lower short- and intermediate-term mortality, while the effect might decline in the long-term. High age, prosthetic valve endocarditis, and Staphylococcus aureus endocarditis were independent predictors for high mortality. Although improvements have occurred over recent years, infective endocarditis is still a high mortality disease.