This study was undertaken to determine whether continuous warm blood cardioplegia (CWBCP) could be acceptable as an alternative method for myocardial preservation in cardiac surgery. Between December 1991 and June 1993, 100 consecutive patients underwent coronary artery bypass surgery. Four patients who received terminal warm blood cardioplegia were excluded in this study. Fourty-eight patients (Group C) served as historical controls, and fourty-eight patients (Group W) from October 1992 to June 1993 served as a prospective, consecutive cohort for statistical comparison. Two groups varied in the types of myocardial protection and cardiopulmonary bypass (CPB) used: intermittent cold blood cardioplegia and moderate hypothermic CPB were used in Group C and CWBCP and normothermic CPB in Group W. The groups had similar number of bypass grafting, aortic cross-clamping time and CPB time. No patients was died. The prevalence of intraoperative cerebral infarction was equal in both groups (4.2%). The incidence of spontaneous defibrillation at cross-clamp removal was higher in Group W (85.4% versus 8.3%; p < 0.01). Less inotrope (dopamine) at 6 hours after operation was required in Group W (3.27 +/- 2.48 versus 4.78 +/- 2.99 micrograms/kg/min; p < 0.01). The intraoperative urgent use of the intraaortic balloon pump was noticeably less prevalent in Group W (0% versus 12.5%; p < 0.05). Group W patients were more likely to be hemodynamically stable after CPB discontinuing. Serum potassium levels during CPB was higher in Group W (max. 5.67 +/- 0.96 versus 4.39 +/- 0.50 mEq/l), so excessive potassium was eliminated using extracorporeal ultrafiltration. The major drawback of CWBCP was that continuous coronary perfusate occasionally obscured anastomosis site.(ABSTRACT TRUNCATED AT 250 WORDS)