The present investigation attempts to correlate flow measurements made intraoperatively in coronary bypass grafts with clinical outcome. A total of 352 consecutive patients undergoing isolated coronary artery surgery underwent hemodynamic assessment of their bypass grafts (328 internal thoracic artery and 582 saphenous vein grafts) at the end of cardiopulmonary bypass (CPB) by using a 8 MHz pulsed Doppler ultrasound flowmeter. The total patient population was divided into three groups of distinct outcome (A: normal, 228 patients; B: complicated, 106 patients; C: poor, 18 patients) on the basis of a combination of the following parameters: difficult weaning from bypass, use of inotropic drugs, reduced left-ventricular stroke work index, myocardial infarction, intraaortic balloon counterpulsation, and death of cardiac origin. Univariate analysis has shown clinical outcome to be influenced by preoperative clinical condition and not by flow in bypass grafts (average flow per graft [ml/min] was 60 +/- 2 [mean +/- SEM] in group A, 58 +/- 3 in group B and 43 +/- 6 in group C: NS by analysis of variance). Multivariate analysis (Fisher linear discriminant analysis) selected only the two following factors leading to normal (group A) or adverse (groups B and C) outcome: unstable angina (p = 0.026) and duration of additional CPB after unclamping the aorta (p < 10(-5). To conclude, clinical outcome was not influenced by flow as measured in well-functioning bypass grafts by pulsed Doppler technique.