Among 377 patients consecutively treated with primary coronary angioplasty for acute myocardial infarction, in-hospital mortality was higher in patients ineligible than in patients eligible for thrombolysis (14.4% vs 7.8%, p <0.05). It remained dismal (75.9%) in patients with cardiogenic shock, but was similar in lytic-eligible patients and in those who were ineligible because of an increased bleeding risk (7.8% vs 7.2%, p = NS), and was zero in patients with nondiagnostic electrocardiograms.