There is a need for a simple clinical measurement that will indicate the extent of myocardial salvage after successful thrombolysis. This study examined whether coronary artery reperfusion reduced the infarct size as assessed electrocardiographically after thrombolytic treatment. The sum of the (sigma) ST segment area in leads showing ST segment elevation in the 12 lead electrocardiogram at presentation was used as an index of potential myocardial injury (initial ischaemic index). The evolved infarct size at 48 h was assessed by a QRS scoring system. Two groups of patients, both admitted with anterior myocardial infarction within 6 h of onset, were studied. Group 1 (n = 35) received analgesia only and group 2 (n = 33) received thrombolytic treatment either by the intracoronary (streptokinase, n = 13) or intravenous route (anistreplase, n = 20). Reperfusion was assessed angiographically. The mean (SD) potential infarct size assessed by the initial ischaemic index was similar in both groups (group 1, sigma ST area = 115 (60) mm2 and group 2 = 126 (77 mm2). The QRS score representing evolved infarct size was significantly lower in the treated group (4.1 (2.5] than in group 1 (7.8 (2.6]. The 95% confidence intervals for QRS scores based on the admission sigma ST area from patients with successful reperfusion were applied to a third set of patients (n = 22) to test the ability of the admission ST area (myocardial injury) to predict the QRS score accurately. While patients with successful reperfusion had significantly lower QRS scores than those who did not (4.5 (3.1) versus 9.3 (3.4)), the wide confidence intervals caused by inter-individual variability precluded an accurate prediction of the QRS score in an individual from the sigma ST area at time of presentation. There was no difference in infarct size in patients treated early (</= 3 h) (QRS score 4.2(2.8)) or later (3-6 h) (4.1(2.1)). This study provides evidence that sequential electrocardiographic changes are reduced in patients with anterior infarction who achieve reperfusion after thrombolytic treatment and that this benefit is shown with treatment given up to six hours after infarct onset. None the less, the relation between the initial ischaemic index and the evolved QRS score has wide confidence intervals, reflecting inter-individual variability, and does not allow the prediction of a QRS score in an individual patient.