In the prethrombolytic era it was found that infarct size and left ventricular ejection fraction could be predicted using the Selvester QRS score. We evaluated whether infarct size and left ventricular ejection fraction could be predicted by the predischarge QRS score in patients who had received reperfusion therapy and whether considering the configuration of the ST segments and T waves would increase the accuracy of these predictions. We evaluated 51 patients with first anterior wall myocardial infarction who had received reperfusion therapy and predischarge resting technetium-99m-sestamibi scan. The electrocardiograms recorded on the same day of the scan were analyzed for the QRS score and were divided into 3 groups: A, isoelectric ST and negative T waves; B, ST elevation (≥0.1 mV) and negative T waves; and C, ST elevation (≥0.1 mV) and positive T waves. Groups A, B, and C included 12, 23, and 16 patients, respectively. The myocardial perfusion defect extent increased from groups A to C (median 21%, 37%, and 43.5% in groups A, B, and C, respectively; p = 0.023). Similarly, left ventricular ejection fraction decreased (44%, 38%, and 34%, respectively; p = 0.042) from groups A to C. Overall, the correlation between the QRS score and the myocardial perfusion defect extent (ρ 0.249; p = 0.08) and ejection fraction (ρ -0.229; p = 0.11) was poor. A statistically significant correlation between myocardial perfusion defect size and QRS score was found only in group A (ρ 0.599, p = 0.04). Among patients with anterior myocardial infarction who received reperfusion therapy, the predischarge QRS score was predictive of infarct size only in those in whom ST elevation resolved completely. In patients with residual ST elevation there was no correlation between QRS score and infarct size. Copyright (C) 2000 Excerpta Medica Inc.