The in-hospital prognosis of patients with acute myocardial infarction complicated by primary ventricular fibrillation has not been satisfactorily defined. We addressed this question by studying patients with primary ventricular fibrillation derived from a large study (11,712 patients) of intravenous streptokinase in the treatment of acute myocardial infarction. Ventricular fibrillation was considered to be primary when it complicated a first myocardial infarction not associated with heart failure or shock and occurred within 48 hours of hospital admission. The 332 patients with primary ventricular fibrillation represented an overall incidence of 2.8 percent. A significant excess of in-hospital deaths was found in the patients with primary ventricular fibrillation as compared with those in the reference group (10.8 percent vs. 5.9 percent; relative risk, 1.94; 95 percent confidence interval, 1.35 to 2.78). Thrombolytic treatment with intravenous streptokinase did not afford protection against primary ventricular fibrillation. We observed that being over 65 years old had a protective effect against primary ventricular fibrillation (relative risk, 0.6; 95 percent confidence interval, 0.45 to 0.80). Our data do not indicate whether primary ventricular fibrillation is simply a marker for patients at increased risk of death or a direct cause of the increase in mortality. Our results do show, however, that primary ventricular fibrillation occurring in a coronary care unit is a negative predictor of short-term survival in patients with acute myocardial infarction.