Between 1990 and 1992, 346 consecutive patients underwent coronary artery bypass procedures. Ninety-eight patients (group A) from 1990 served as historical controls, and 248 patients (group B) from 1991 to 1992 served as a prospective, consecutive cohort for statistical comparison. The two groups varied in the type of myocardial protection used: intermittent cold crystalloid cardioplegia was used in group A and continuous warm blood cardioplegia in group B. (Two patients in group A received intermittent cold blood cardioplegia, and these 2 patients are grouped with the crystalloid group for the sake of convenience. The presence or absence of these 2 patients did not alter the group A statistics in any noticeable manner.) Class IV high-risk groups demonstrated a 63% reduction in mortality ( p = 0.07), and overall group B experienced a 28% reduction in mortality (4.4% versus 6.1%; p = not significant), an 86% reduction in perioperative myocardial infarction rate (1.6% versus 12.2%; p < 0.05), a 20% reduction in postoperative bleeding (275 versus 345 mL · day −1 · m −2), and a marked reduction in reentry rates ( p = 0.05). Also noted was a 32% reduction in postoperative ventilation requirements (25 versus 37 hours; p = 0.05). Less inotrope was required and intraoperative stroke was not seen in the patients with warm blood cardioplegia. Group B patients were less likely to have development of complex postoperative arrhythmias. Ventricular fibrillation at unclamping was noticeably rare (2.0% in group B versus 84% in group A; p < 0.05). The average group B heart resumed sinus rhythm 72 seconds after declamping. Creatine kinase MB fractions were 7.4% in group B versus 8.2% in group A ( p = 0.07). The average postoperative cardiac index was 16% higher in group B patients (3.0 versus 2.5 L · min −1 · m −2; p = 0.05), and low output syndrome was reduced by 22% ( p = not significant) in group B patients. Cold phrenic nerve injury did not develop in any patient in group B. Coronary sinus perforation occurred once but was easily repaired. Normothermic continuous blood cardioplegia is superior to other forms of cardioplegia and is our choice for myocardial protection.