BACKGROUND: Even with the current aerobic techniques in myocardial protection, perioperative myocardial infarction can still occur. In this study, we sought to find out whether there is a method-specific mechanism of ischemic injury in coronary bypass surgery using an on-pump beating heart technique. For this reason, we investigated localization and the extent of myocardial infarction in correlation to the severity of coronary artery stenosis. We discuss strategies for reduction of infarction incidence. METHODS: 34 patients, who after isolated coronary bypass procedure developed non-Q or Q wave infarction were selected for the study. In 12 of 34 patients (group A) beating heart technique and in 22 of 34 patients (group B) cardioplegic arrest was used for myocardial protection. The study was conducted retrospectively and included patients with stable, unstable, and postinfarction angina without preoperative enzymatic evidence of ischemic injury and with technically noncomplicated coronary bypass grafting. Excluded from the study were emergency and redo procedures. In group A, in all instances, the left anterior descending artery was grafted as the last vessel and while distal anastomoses were constructed, coronary perfusion was maintained by using a perfusion catheter. RESULTS: Most patients in both groups had triple-vessel disease (10 of 12 patients group A; 17 of 22 patients group B), all patients received complete revascularization; 7 of 12 patients in group A and 5 of 22 patients in group B had significant or critical left main stem stenosis. In group A, 11 infarctions occurred in LAD-(12 stenotic), 1 in CX-(11 stenotic), and 1 in RCA-(11 stenotic) supply area, four were Q wave infarctions. In group B, eight infarctions occurred in LAD (22 stenotic), four in CX (17 stenotic), and 14 in RCA (20 stenotic) supply area, eight were Q wave infarctions. In group A, the infarction incidence in the LAD area was 10-times higher than in CX and RCA areas. In group B, the infarction incidence in the RCA area was 2- and 3-times higher than in the LAD and CX areas, respectively. In both groups no correlation between infarction incidence and severity of stenosis was observed. CONCLUSIONS: Using an on-pump beating heart technique, higher coronary perfusion pressures, avoidance of extreme upward retraction of the heart during revascularization of the CX-branch, as well as choosing the revascularization of the LAD as the first vessel, could possibly contribute to better myocardial protection. In hearts arrested with cardioplegic solution, the right ventricle is probably more susceptible to ischemic injury, especially when RCA is poorly collateralized. For adequate protection, choosing the revascularization of the RCA as the first vessel with immediate repeated cardioplegia via a RCA graft, higher perfusion pressures and antegrade with retrograde cardioplegia delivery, may be advantageous.