IN A PREVIOUS issue of this journa1,l the details of our national cooperative ControIled study of long-term anticoagulant therapy in cerebral ischemia and cerebral infarction were presented along with an interim report covering some eighteen months. At that time, longterm anticoagulation did not appear to reduce the mortality and there was an added risk due to hemorrhagic complications. The study, embracing 7 clinical centers and extending over a period of forty-two months, was brought to an end in August 1961, and this is a report of our findings and final conclusions. The important aspects of the study, including selection and classification of cases and method of anticoagulation, were fully covered previously and will not be repeated here. A total of 443 patients were enrolled in the study, each patient being assigned to the anticoagulant or nonanticoagulant therapeutic groups on a statistically random basis by means of sealed envelopes. All the patients in the study were divided into 5 diagnostic categories: (1) transient ischemic attacks, (2) thrombosis-in-evolution, (3) thrombosis-completed stroke, ( 4 ) thorem (thrombosis or embolism), and (5) cerebral embolism. Analysis of the control and treated groups showed that they were remarkably comparable in most respects (table 1). The number of cases involving the internal carotidmiddle cerebral and vertebral-basilar systems varied from category to category; in the whole series there were 328 in the carotid and 102 in the vertebral-basilar. In regard to blood pressure, 59.6 per cent of the entire group had an elevated blood pressure. To determine the effect of anticoagulant therapy, the cases were analyzed as to the incidence of (1) deaths, (2) progression of cerebral infarction due to either extension of the infarcted area or a new infarction, (3) the advent of extracerebral thrombotic episodes (ECTE) , including myocardial infarction, pulmonary embolism, thrombophlebitis, and so forth, and (4) hemorrhagic complications. Patients whose anticoagulants were stopped within two weeks of death when the patient's condition was deteriorating were reckoned as deaths while on anticoagulants.