In the late Eighties, a series of observations from several institutions around the world has dramatically revolutionized the traditional notion that the occurrence of myocardial infarction was related most of the time to the development of critical flow-limiting coronary stenosis. All these studies showed that the infarct-related artery had only minimal or mild stenosis in about two thirds of the cases. Therefore, contrary to our previous beliefs, in clinical practice the detection of coronary stenosis has a lesser role in the prognostic assessment of patients with ischaemic heart disease, unless associated with extensive ischemia or with phases of instability. In fact, the major determinants of prognosis are represented by age, left ventricular function, effort tolerance and especially by the clinical stability or instability of angina. According to the Bayesian an theorem, in low risk patients any diagnostic test has a very low predictive accuracy, unless very high specificity criteria are used. The value of diagnostic tests in the assessment of patients' prognosis should be evaluated in intermediate risk groups. The emphasis of clinical research has, therefore, shifted from the detection of flow-limiting stenosis to the study of the multiple and varied dynamic causes of stable and unstable ischaemia, where the possibilities of making new, seminal observations are greater.