Between 1977 and 1992 a total of 163 consecutive patients underwent emergency coronary artery bypass grafting after acute coronary occlusion (94% after failed angioplasty). Patients were divided into four groups according to the method used for myocardial protection. The crystalloid cardioplegia group included 30 patients operated on from 1977 to 1980; the hypothermic fibrillation group included 60 patients (1980 to 1986); the blood cardioplegia group included 36 patients (1986 to 1989); and the blood cardioplegia with controlled reperfusion group included 37 patients (1989 to 1992). Preoperative data, ischemic time interval, collateral blood flow, intraoperative data, regional wall motion, global ejection fraction, myocardial infarct-specific electrocardiographic changes, enzyme release, rhythm disturbances, mortality, prevalence of intraaortic balloon pumping, and inotropic support were assessed in this retrospective study. Our data indicate that the current spectrum of patients undergoing emergency coronary artery bypass grafting after acute coronary occlusion are at a significantly higher risk compared with those 15 years ago, that is, increase in age (53 +/- 1 versus 59 +/- 2 years; p < 0.05), three-vessel disease (38% versus 3%; p = 0.004), acute occlusion of the left main coronary artery (11% versus 0%; p = 0.02), preoperative cardiogenic shock (35% versus 3%; p = 0.007), prevalence of acute two-vessel occlusion (22% versus 3%; p = 0.05), prevalence of previous infarction (59% versus 23%; p = 0.04), and duration of ischemia (3.0 +/- 0.2 versus 4.1 +/- 0.3 hours; p < 0.05). Despite the increase in patients with severely compromised ventricular function during recent years, the overall hospital mortality decreased to 5% (2/37) when maximal protection of the ischemic and remote myocardium was performed (preoperative intraaortic balloon pump, combined antegrade/retrograde substrate-enriched blood cardioplegia, warm induction, controlled reperfusion, prolonged vented bypass). Single-vessel disease was always associated with a low mortality, whereas mortality could be reduced with controlled blood cardioplegia in patients with multivessel disease (6%) and cardiogenic shock (15%). The immediate return of regional contractility in the previously ischemic area after controlled reperfusion might serve as an explanation for these favorable results. After unmodified blood reperfusion, normokinesis or slight hypokinesis occurs in only 34% to 46% in the early postoperative period (1 to 4 weeks) in comparison with 86% after controlled blood cardioplegia reperfusion (p < 0.05). We conclude that there is a significant increase in risk factors in patients undergoing emergency coronary artery bypass grafting and that improved methods of intraoperative myocardial protection are needed for these compromised patients.