The clinical relevance of reciprocal changes in the ST segment occurring at the time of acute myocardial infarction was studied prospectively in 85 consecutive uncomplicated cases. Reciprocal depression of the ST segment was defined as depression of 1 mm or more in electrocardiogram leads other than those reflecting the infarct. All patients underwent maximal, symptom limited treadmill stress testing two weeks after the infarct and coronary angiography six weeks after infarction. Forty six patients had inferior, 34 anterior, and five true posterior infarction. Of the 51 patients with reciprocal changes, 45 (88%) developed exercise induced ST segment depression in areas remote from the infarction zone. At angiography all 45 patients were shown to have stenoses greater than 70% in at least two major vessels. Four patients had negative exercise electrocardiograms and were sequently shown to have single vessel disease subtending their infarct, and the remaining two patients had a false negative treadmill test result. Of the 27 patients without reciprocal changes, 21 (78%) had negative treadmill stress test results associated with single vessel coronary disease. Five had positive stress test results and multivessel coronary disease, and one had a false negative stress test result. The remaining seven patients had ST segment elevation without Q wave formation in the reciprocal areas and were assessed separately. Of these, six had positive stress test results and multivessel coronary disease and one had a negative stress test result and single vessel coronary disease to the infarct area. Twenty one patients with anterior infarcts (62%) and 27 with inferior infarcts (59%) had reciprocal changes. No differences emerged in the relation between infarct site, reciprocal change, and presence of additional coronary disease. At follow up of the 51 patients with reciprocal changes in the ST segment 36 had become symptomatic, of whom 29 had undergone coronary artery bypass surgery. By contrast, only four of the 27 patients without reciprocal changes in the ST segment had developed symptoms, and two of these had undergone coronary revascularisation. Reciprocal ST segment depression at the time of acute myocardial infarction may identify patients with severe coronary disease who are at risk of subsequent cardiac events and appears to be as reliable as results of early postinfarction treadmill stress testing in predicting the underlying coronary anatomy. When the electrocardiogram does not show reciprocal changes treadmill testing provides valuable additional information.