The incidence and prognosis of myocardial infarct extension were studied prospectively among patients presenting without symptomatic heart failure (Killip class I or II). Infarct extension was defined as a new clinical event (recurrent pain, arrhythmias or worsening hemodynamic status) that occurred at least 24 hours after a documented myocardial infarction and accompanied by at least two of the following: 1) new QRS changes on the scalar ECG, 2) reappearance of the MB band of creatine kinase (MB CK), or 3) a new peak of total CK. Suspected infarct extension was defined as a new clinical event accompanied by only one of the above. Survival rates were calculated by the life-table method and compared at 1 year by chi-square analysis. Follow-up information was 99.4% complete. Between January 1971 and April 1977 there were 458 admissions for myocardial infarction and 58 episodes of definite or suspected extension (13%). Hospital mortality in patients with extensions was 36% (21 of 58) vs 9% (36 of 400) in those without extensions (p<0.001). Of patients in both groups who survived hospitalization, 1-year survival was 76% in patients with definite or suspected extension and 91% in those without extension. The incidence of infarct extension in this group of patients was lower than that in other reports. The poor hospital and 1-year survival rates mandate early therapeutic intervention in these patients.