A satisfactory definition of unstable angina continues to be elusive.Like stable angina, it is a clinicalsyndrome, and mustbe defined in those terms. In both cases, it is understood that myocardial ischaemiabut not infarction is responsible for the symptoms. For angina to earn the “unstable” label, theremust have been the recent development or deterioration of symptoms. Traditionally, “recent” has meant within the last month, but it has become increasingly clear thatthe time frame is of critical importance - the patient with a sudden irruption of severe chest pain in the last two days is likely to have a different pathology and prognosis from the individual who first developed exercise-induced pain two weeks ago, which has not worsened during this period. Likewise, the patient who hadhis last attack of pain one week agois very different from one who had his last attack one hour ago; indeed in the HINT study, there was a high incidence of myocardial infarction undetected on admission in the lattergroup (Br Heart J 1986;56:400-13).A problem that has hardly been addressed is “When does unstable angina stop being unstable?”. The answer must be when it either becomes stable (i.e. stops getting worse),or proceeds to myocardial infarctionor death.Most studies have shown that the vast majority of patients stablise quickly, often losing their symptoms completely. There is, however, a group of patients who fail to respond promptly to medical treatment, and it isthese patients who are most likely to go on to angiography, angioscopy, angioplasty, surgery, myocardial infarction or death. Thus, the enormously valuable information we have obtained from, for example, angiography and necropsy studies applies essentially only to this subset of unstableangina patients, albeit they are themost severe. Unfortunately, such studies tell us relatively little aboutthe more dynamic aspects of the disorder, such as the role of coronary vasomotion.It is only relatively recently that clinicians have appreciated that they have quite simple means of suspecting the mechanisms involved. Thus, the history of progressive exercise-induced angina on the one hand or angina only at rest on theother must tell us something of the underlying physiopathology. Likewise,the fact that the symptoms respond to rest, or beta-blockers, or calcium antagonists also provides evidenceas to causation.Thus, while all would agree that angina is “unstable” whenit has recently developed or worsened, it must bediscussed in subsetsdefined by their history and response to treatment.In this way, we may be able to assign a patient to a particular clinicalgroup, which will indicate that it is likely that he has a particular physiopathology, which inturn will suggest the most appropriate management.