Seventeen of 34 consecutive patients undergoing coronary artery bypass grafting were randomly assigned to one of two methods of myocardial preservation. With the cold cardioplegic method (Group A), a 4 degrees C. asanguineous solution with 30 mEq. of potassium per liter was infused into the aortic root for about 2 minutes immediately after aortic cross-clamping and again after about 45 minutes or when myocardial temperature rose above 19 degrees C. External cardiac cooling was provided by constant infusion of 4 degrees C. Ringer's solution into the pericardium. Seventeen patients were assigned to simple cardiac cooling by hypothermic systemic perfusion before aortic cross-clamping plus external cardiac cooling (Group B). Electromechanical activity ceased within 1 to 2 minutes in Group A but continued throughout the ischemic period in 14 patients in Group B. Myocardial temperature (mean for all observations) during aortic cross-clamping was 17.2 +/- 0.44 degrees C. In Group A and 24.0 +/- 0.70 degrees C. in Group B. Operating conditions were better in Group A. Card-ac function early postoperatively was good in both groups clinically and according to measurements, but only in the cold cardioplegic group (A) was cardiac index not adversely affected by longer cross-clamp time. Myocardial necrosis occurred in both groups but was probably less in the cold cardioplegic group. Thirteen patients (76 percent) in Group A had no electrocardiographic evidence of myocardial injury, compared with eight (47 percent) in Group B (p = 0.08). Eleven (65 percent of Group A had no or short-lived appearance of ceatine phosphokinase isoenzyme (CK-MB), compared with six (35 percent) of Group B (p = 0.08). Time-related CK-MB and SGOT mean levels were consistently lower in Group A.