The accuracy of studies on the prognosis of angina pectoris depends on two factors. In the first place, errors will naturally occur because of mistaken diagnoses. There are a great many types of distress or pain in the chest that are not angianl in character which are often confused with true angina pectoris. Second, the duration of life after the symptoms have developed will obviously depend on the care that is taken in ascertaining when the disease actually first manifested itself. Frequently patients date the onset at a time when an attack of severe pain in the chest occurred, but on questioning them directly as to the first time that they had this sort of distress, even in a mild form or on walking, one learns that the disease began 1 or more years previously. The duration of life after the onset of the disease would consequently be greater according