Infective endocarditis is a severe disease with high mortality, and results most often from the combination of bacteraemia (sometimes provoked) and a predisposing cardiac condition. Prophylaxis for infective endocarditis has been recommended by different countries on the basis of the supposed pathophysiology of the disease, although no randomised clinical trial has confirmed its efficacy. We review the data presented over the past few decades, challenge the principles underlying prophylaxis recommendations, and analyse the arguments that explain the general tendency in very recent years to decrease prophylaxis indications. Such arguments include the probable important role of everyday-life bacteraemia in the occurrence of infective endocarditis, the estimated huge number of prophylaxis doses to be given to avoid a single case of infective endocarditis, and the lack of scientific evidence to identify those procedures that should lead to prophylaxis. Recommendations for prophylaxis are now essentially focused on patients with high-risk predisposing cardiac conditions before dental procedures.