A prospective, randomized trial was instituted to determine whether blood cardioplegia (BC) could reduce the morbidity and mortality for patients undergoing urgent coronary bypass for unstable angina. One hundred forty patients who came to the hospital with prolonged angina at rest and who required urgent revascularization because their symptoms were resistant to medical therapy were randomized to receive BC (n = 70) or crystalloid cardioplegia (CC) (n = 70). The operative mortality rate was 2.8%, the incidence of myocardial infarction was 8.6%, the incidence of low output syndrome was 18% and morbidity (myocardial infraction or low output syndrome) was 23%. Patients who received BC had a significantly lower mortality rate (BC, 0%; CC, 5%; p less than 0.05) and incidence of myocardial infarction (BC, 4%; CC, 13.5%; p less than 0.05) or low output syndrome (BC, 10%; CC, 19%; p less than 0.05). The highest postoperative CK-MB level was less after BC (BC, 31 +/- 17 U/L; CC, 56 +/- 13 U/L; p less than 0.05). Preoperative predictors of postoperative morbidity in addition to the type of cardioplegia included the response to medical therapy, persistent ischemic electrocardiographic changes, left ventricular ejection fraction, and age. A multivariate analysis identified the type of cardioplegic protection (p = 0.008) and age (p = 0.05) as significant independent predictors of postoperative morbidity. BC reduced the risk of urgent revascularization for unstable angina.