For a long time it has been known that bacteraemias caused by medical or dental procedures may cause endocarditis in patients with specific types of congenital or acquired heart disease. In the 1940s it was thought that the administration of antibiotics before such procedures would prevent endocarditis. However, the beneficial effect of this preventive measure on the incidence of endocarditis did not live up to its expectations. Quite soon it became obvious that prophylaxis was not 100% efficacious in man, although it did prevent endocarditis in animals. A controlled study into the protective effect of prophylaxis in humans has never been carried out. In the last decade it has become dear from case-control studies that endocarditis prophylaxis is not a very effective preventive measure but that it reduces an already small risk even further. In this article the theoretical background of endocarditis prophylaxis and possible explanations for its lack of effect are discussed.