Background Several attempts from single institutions to treat acute myocardial infarctions with bypass surgery never reached widespread acceptance in the cardiology and surgical community. Owing to a variety of new surgical techniques, this old dogma has to be reconsidered under the light of patient-adjusted optimal treatment algorithms. Methods Between August 2002 and August 2007, 112 patients, mean age of 66 years (range, 41 to 85 years), underwent emergency coronary artery bypass grafting (untreatable or rejected by the referring cardiologists within 48 hours after onset of symptoms). Thirty-seven patients (33%) exhibited cardiogenic shock, and 18 (16%) had prior cardiopulmonary resuscitation. Preoperative support by intraaortic balloon pump was initiated in only 10%, and 65% had left main stem stenosis. Results All patients showed a significant elevation of cardiac markers (creatine kinase-MB) and ST-segment elevation. The mean number of grafts was 2.4 (range, 1 to 4). The cardiopulmonary bypass time ranged from 48 to 261 minutes. Intraaortic balloon pump for weaning from extracorporeal circulation was used in 42 patients (38%); 3 patients needed extracorporeal membrane oxygenation support. Postoperative complications included rethoracotomy for bleeding in 4% and stroke in 2%. Thirty-day mortality was 20% in the whole group, 30% in the group with cardiogenic shock, and 15% in those without hemodynamic deterioration (p = 0.044). The multivariate analysis revealed the preoperative need for catecholamines as the only risk factor for 30-day mortality (odds ratio, 6.4; 95% confidence interval, 2 to 21; p = 0.002). Conclusions Emergency coronary artery bypass grafting in patients with acute myocardial infarction can be performed with acceptable results, especially in those without cardiogenic shock. Therefore, operative revascularization should not be considered only as a rescue therapy.