The effect of three cardioplegic protocols on perioperative myocardial injury was studied in 62 coronary artery bypass grafting (CABG) patients randomized into three groups with either antegrade or retrograde cold blood cardioplegia, or coronary sinus occlusion during antegrade supply. During the aortic cross-clamp time anterior and posterior septal temperatures were recorded, indicating the distribution of cardioplegic solution within the myocardium. Serum creatine kinase (CK), CK-isoenzyme MB and myoglobin as well as 12-lead electrocardiograms (ECG) were analyzed. Statistical analysis showed no effect of the cardioplegic protocol, whereas the patient's preoperative status, aortic cross-clamp time and intraoperative myocardial temperature had significant (P < 0.05) effects on immediate postoperative CK and CK-MB enzyme release. Creatine kinase-MB peak values were significantly increased in patients with major vessel disease and reduced left ventricular function (92 +/- 53 U/l versus 67 +/- 25 U/l). Both CK and CK-MB values were significantly higher in patients with aortic cross-clamp times of more than 1 h than in patients with shorter clamping times (661 +/- 188 and 78 +/- 40 U/l versus 500 +/- 200 and 57 +/- 24 U/l). Patients with 22 +/- 3 degrees C myocardial temperature before terminal cardioplegia had significantly elevated CK as compared to patients with temperatures of 15 +/- 2 degrees C (665 +/- 185 U/l versus 510 +/- 211 U/l). However, enzyme peak values had only poor predictive power for postoperative ECG changes, suggesting that enzyme peaks were not necessarily a sign of perioperative ischemia. Patients with major vessel disease and reduced myocardial function, with aortic cross-clamp time of more than 1 h and/or inadequate intraoperative myocardial cooling may be highly susceptible to global ischemia and operative procedures, and therefore show elevated peak enzyme levels shortly after surgery. In contrast, elevated myoglobin peaks within 1 h after aortic declamping were significantly correlated to perioperative signs of transient ischemia (P < 0.02).