Prospective, randomised trials are the key to evidence-based medicine. They have brought considerable progress to the field of cardiology. They have, however, methodological constraints that result in the inclusion of only a small proportion of patients presenting with a given disease. In contrast, registries include all patients, without any selection, and therefore reflect everyday practice in the "real world". They are the key to epidemiological data, but they can also be used to tackle questions or issues that are unresolved by randomised trials. Both approaches are complementary and should be used to define the optimal management of patients. Lessons derived from registry data in acute coronary syndromes in the recent past in France are presented.