Prognosis constitutes the possible outcomes of a disease and the frequency with which they can be expected to occur. The need for prognostic determination in cardiology may arise from the patient's questioning ("What is going to happen to me?"), or from the cardiologist himself trying to decide on therapy (what should be done with an NYHA class IV patient with congestive heart failure refractory to treatment, if he has a cardiac arrest?), or on screening (what is the clinical evolution of an asymptomatic 5-cm diameter abdominal aortic aneurysm?). Prognosis is also important in making inferences about quality of care, when comparing outcomes in populations from different contexts and institutions. In order to establish an accurate prognosis in a specific patient, one has to examine the clinical evolution of groups of patients with the same disease (stroke, for example), refining this process by inclusion of the patient in specific subgroups defined by, for example, age (young patients fare better than old ones), clinical variables (hemorrhagic versus thrombotic stroke) or existing comorbidities (hypertensive patients have a worse outcome). In this article we discuss the guidelines for critical appraisal of a prognostic study, in terms of the validity, importance and applicability of its results.