An appreciation of the natural history of angina pectoris is important when deciding on the place of new and potentially dangerous forms of treatment. During 1950-1975, 268 patients with angina were diagnosed and followed up in my London general practice. The annual incidence, in adults over 40, was five per 1,000 and increased with age.During the period of follow-up, half the patients died, an annual mortality of 4.6 per cent. However, among the survivors one third ceased to suffer anginal symptoms spontaneously and without specific therapy. Of those who continued to suffer from angina, in 71 per cent the condition was graded as minor, in 27 per cent as moderate, and in only two per cent were the attacks severe and disabling. Usually the angina was primary (77 per cent) and it was secondary, after myocardial infarction, in 23 per cent.Of the 134 deaths, three quarters were from a cardiovascular cause. This group of angina patients had a 2:1 times greater observed, than expected, risk of dying (O/E ratio). The O/E mortality ratio fell progressively with age. It was highest in the 40-49 decade (4.0) and lowest in the over 80s, when the observed mortality rate was less than expected (0.9). The O/E mortality ratio was higher in men (2.3) than in women (1.7).From this survey I conclude that angina does not have a uniformly bad prognosis, and that with the advent of beta-adrenergic blockers, the proportion considered for angiocardiography and aorto-coronary bypass grafting should be less than five per cent of all patients with angina.