To assess the predictability of need for emergency surgery after coronary balloon angioplasty.Nonrandomized intervention study.Nonprofit university hospital.Prior to balloon angioplasty, 1000 consecutive patients were assigned to either the "standby" group (189 patients [19%]) or the "no-standby" group (811 patients [81%]). Patients in the standby group (intervention coordinated with cardiac surgery) included all operable patients undergoing angioplasty of their largest coronary arteries that were not currently or previously totally occluded or collateralized; the no-standby group consisted of the remainder of patients.Allocation to coronary angioplasty with or without surgical standby.Need for bypass surgery, occurrence of myocardial infarction, and mortality from complications of angioplasty.Bypass surgery immediately after angioplasty was done in one patient in each group (standby, 0.5%, vs no-standby, 0.1%). The frequency of infarction was 5% vs 4%, respectively. All eight deaths occurred in the no-standby group (1.0%), but none of them were consequences of a lack of surgical standby. They occurred in situations in which bypass surgery would not have changed the outcome (two cardiac failures late after technically successful angioplasty for postinfarct cardiogenic shock, one in-laboratory rupture of an unrecognized ventricular pseudoaneurysm, and one protamine reaction), secondary to acute problems late after successful angioplasty (two sudden deaths and one vessel occlusion in an inoperable patient), or despite surgery (one patient with left main stem dissection).Performing roughly 80% of coronary angioplasties without surgical standby did not increase patient risk. Coronary angioplasty without surgical backup, albeit not an ideal setting, appears ethically feasible in selected patients if dictated by logistic considerations.