The various factors influencing the result of treatment have been studied in a series of 100 consecutive patients undergoing aorto-coronary bypass graft surgery. There were three operative deaths and twelve cases of post-operative infarction. Longterm, 85% of them were three year survivors, clinical improvement being maintained in 84% of them. Post-operative tests showed that 70% of patients have no pain on the maximal exercise test, but 48% had ischaemic depression of the ST segment. From among the 47 patients who had follow-up arteriography, 78% of the grafts were patent, but no improvement of the contractility of the left ventricle could be demonstrated (pre-operative ejection fraction 47 +/- 3%, post-operative 49+/- 3%). The pre-operative clinical features rarely give any indication of the prognosis. The same is true in the present series of the condition of the left ventricle, bearing in mind the fact that patients with grossly impaired ventricular contractility (ejection fraction below 30%) were routinely excluded from surgery. It is essentially the state of the coronary arterial network, as assessed by a score made up from the degree and number of stenoses, the quality of the distal bed, and the anatomical distribution, which will pick out those patients more at risk both from the surgery and from post-operative death and in whom the longterm result is likely to be disappointing with no improvement in function, with obstruction of the grafts, or with secondary death.