This study attempts to relate flow findings in internal mammary (IMA) and saphenous vein coronary artery bypass grafts to postoperative outcome. From 262 patients undergoing coronary artery bypass grafting, 601 electromagnetic flow measurements were obtained in IMA and saphenous vein grafts, and free graft flow was measured in 227 IMAs prior to grafting. Retrograde flushing of the IMA with diluted papaverine hydrochloride resulted in a marked increase in IMA free flow (124 +/- 4 mL/min versus 66 +/- 5 mL/min; p < 0.001). However, IMA free flow did not correlate with electromagnetic flow measurements after grafting to the left anterior descending coronary artery. The use of IMAs with free flows lower than 50 mL/min did not affect clinical outcome. Flow measured in saphenous vein grafts (66 +/- 9 mL/min) with an electromagnetic flowmeter was significantly greater (p < 0.001) than that in the IMA grafted on the left anterior descending coronary artery (36 +/- 3 mL/min) under comparable hemodynamic conditions. For the purpose of data analysis, patients were separated into three groups based on increasing incidence of complications: levels 0, 1, and 2. Patients with an uneventful outcome had a mean graft flow at chest closure of 51 +/- 3 mL/min versus 51 +/- 4 mL/min for patients in complication level 1 and 45 +/- 11 mL/min for patients in complication level 2 (p = not significant). Free flow measured in a vasodilated IMA was a poor predictor of flow into a grafted IMA and did not affect clinical outcome. We were unable to validate any flow limit to use of the IMA.(ABSTRACT TRUNCATED AT 250 WORDS)