We assessed predicting final infarct size (using predischarge Selvester score) by 3 electrocardiographic variables in 267 patients with first anterior wall acute myocardial infarction (AMI) undergoing (n = 86) or not undergoing (n = 181) thrombolysis. Patients with previous AMI or inverted T waves in leads with ST elevation were excluded. The sum (sigma) of ST elevation, the number of leads with ST elevation, and the initial electrocardiographic pattern were determined on the admission electrocardiogram (absence (QRS-) or presence (QRS+) of distortion of the terminal portion of the QRS in > or =2 leads (J point > or =0.5 of the R-wave amplitude in leads I, aVL, V4 to V6, or presence of ST elevation without S waves in leads V1 to V3). There was no association between sigmaST elevation and final infarct size in patients who did or did not receive thrombolytic therapy. Analysis of covariance showed that the number of leads with ST elevation (F = 19.6), thrombolysis (F = 25.2), and QRS+ initial pattern (F = 19.5) were all associated with final infarct size (p <0.0001 for all). Among patients who did not receive thrombolytic therapy, the average Selvester score was 19.7+/-9.9 for the QRS- patients and 26.1+/-10.4 for the QRS+ patients (p = 0.02). Among patients who received thrombolytic therapy, the average Selvester score was 11.7+/-9.8 for the QRS- patients and 24.2+/-10.1 for the QRS+ patients (p <0.0001). Thrombolysis reduced final Selvester score only in the QRS- group (p <0.00001), but not in the QRS+ group (p = 0.45). It is concluded that (1) final Selvester score in anterior wall AMI can be predicted by the number of leads with ST elevation, the initial electrocardiographic pattern, and thrombolysis, and (2) thrombolysis reduces final Selvester score only in patients with QRS- pattern.