The effectiveness of diagnostic, preventive and therapeutic procedures, whose efficacy has been assessed in clinical trials, should be tested in a real treatment scenario. The procedures used in acute myocardial infarction (AMI) management can be evaluated by means of cohort studies that include all consecutive patients admitted to one or several hospitals. Such studies are called hospital registries. They are simpler to organize and cheaper than clinical trials. On the other hand, the AMI population-based registries allow the establishment of the incidence and mortality rates, as well as case-fatality as they include those patients who die before reaching hospital facilities. In both types of registries a set of variables on co-morbidity, age, sex, severity, and the utilization of procedures along with the course of the disease are systematically recorded in each patient using standard definitions to warrant the internal validity. In hospital registries, the external validity of the results will depend on whether the sample of hospitals represents the population where it was obtained. A good registry should include patients with a wide age range, allow the analysis of specific subgroups of patients such as non-Q wave or first AMI to allow for comparison with other registries. In addition, it should also permit a mid-term follow-up, respect ethical issues, receive appropriate funding and keep a multidisciplinary team involved in its design and development.