Leukocyte depletion at reperfusion may have a role in myocardial protection when combined with terminal cardioplegia. We applied this method in a selected group of 68 patients with coronary artery bypass grafting either for elective surgical procedures (n = 38) or emergency surgical procedures with the use of a preoperative intraaortic balloon pump (n = 30) because of developing acute myocardial infarction. Basic cold potassium crystalloid cardioplegic solution was used. During delivery of leukocyte-depleted terminal cardioplegic solution, warm arterial blood delivered from cardiopulmonary bypass was passed through a leukocyte removal filter, mixed with potassium crystalloid cardioplegic solution, and administered to the aortic root for the first 10 minutes of reperfusion. Patients were randomized into three groups for reperfusion: whole blood, terminal cardioplegic solution, and leukocyte-depleted terminal cardioplegic solution reperfusion groups. In elective coronary artery bypass grafting, no significant difference was found in the clinical data. However, in emergency coronary artery bypass grafting, the leukocyte-depleted terminal cardioplegic solution group (n = 10) showed significantly lower peak creatine kinase MB levels (leukocyte-depleted terminal cardioplegic solution versus terminal cardioplegic solution versus whole blood: 27 +/- 11, 56 +/- 13, 74 +/- 18, respectively; p < 0.05) and maximum dopamine doses required at the weaning of cardiopulmonary bypass (6.3 +/- 1.1 versus 11.2 +/- 3.3 versus 9.2 +/- 2.2; p < 0.05) than did the terminal cardioplegic solution (n = 10) and whole blood groups (n = 10). Moreover, the leukocyte-depleted terminal cardioplegic solution group showed significantly lower difference of malondialdehyde between arterial and coronary sinus blood (0.15 +/- 0.09 versus 0.36 +/- 0.06 versus 0.06 +/- 0.12 nmol/ml, p < 0.05) than did the terminal cardioplegic solution or whole blood groups. These results showed that leukocyte-depleted terminal blood cardioplegic solution may have a role in attenuating reperfusion injury in patients with critical conditions such as preoperative myocardial ischemic injury.