Relief of angina by nitroglycerine has been attributed to two possible mechanisms: (1) Increase in myocardial blood supply by direct action of the drug on the coronary arteries. (2) Reduction in myocardial oxygen demand by action of the drug on the systemic circulation. Sublingual application of nitroglycerin in previous studies did not allow the ruling out of the operation of either or both mechanisms. This study was, therefore, designed in such a way that the effect of the direct action of nitroglycerin on the coronary bed on angina could be studied in the absence of changes in the systemic circulation. In 25 patients undergoing cardiac catheterization and coronary arteriography as possible candidates for revascularization surgery, nitroglycerin, 0.075 mg in 1 ml of 5% dextrose, was injected into the left coronary artery through the angiographic catheter during angina pectoris induced by pacing. Coronary sinus blood flow by the continuous thermodilution method, femoral artery blood pressure, and lead V5were recorded continuously. In 20 patients the procedure was repeated with injection into the right coronary artery. In none of the 25 patients did the intracoronary injection of nitroglycerin alleviate the angina during the 1-min observation period, whether injected into the obstructed artery or into the artery supplying collaterals to the obstructed artery. The intracoronary injection was ineffective despite a significant increase in coronary sinus blood flow in 14, lasting 26 sec on the average. This suggests that the increase in blood flow did not occur in the ischemic areas where it was needed, but in other areas where the arterioles were not maximally dilated by ischemia. This explanation is further supported by the fact that in five patients with very severe restriction of the left coronary artery system the coronary sinus blood flow failed to increase in response to intracoronary nitroglycerin during angina, but increased significantly in response to nitroglycerin after discontinuation of pacing and disappearance of angina. In six patients, 0.2 mg nitroglycerin injected intravenously 1 min following the intracoronary injection relieved the angina unaffected by the preceding intracoronary injection. The relief was associated with a fall in arterial blood pressure and coronary sinus blood flow. The study indicates that the direct action of nitroglycerin on the coronary bed plays little, if any, role in the antianginal effect of the drug, which appears to be due entirely to the action of the drug on the systemic circulation.