To examine the prognostic value of early radionuclide imaging in patients with transmural acute myocardial infarction, 222 patients in Killip class I and II were studied prospectively within 24 hours of the onset of symptoms. The 30-day mortality rate for the entire group was 11% (25 of 222). Univariate analysis indicated that an initial radionuclide left ventricular ejection fraction (EF) of less than 0.30 was associated with the greatest relative risk (RR = 6.6), although the percent of abnormally contracting regions (RR = 3.9) and thallium-201 defect index (RR = 3.3) were also significant risk factors. Stepwise logistic regression indicated that addition of EF resulted in the greatest improvement over the best clinical model (Killip class and chest radiographic findings) for the prediction of 30-day mortality (chi 2 improvement = 12.8, p less than 0.0005). Using the optimal model for prediction of mortality (EF and Killip class), a high-risk group with a 30-day mortality rate of 39% (90-day mortality 47%) and a low-risk group with a 30-day mortality rate of 3% (90-day mortality 4%) was identified. In clinically stable patients with transmural acute myocardial infarction, early assessment of EF in conjunction with clinical evaluation, is a valuable method for early identification of high-risk subsets.