The ability of exercise ECG body surface mapping to detect the presence and distribution of coronary artery disease was investigated in 76 patients presenting with chest pain. The ECG data were recorded from 16 leads regularly placed over the left praecordium. All 16 leads were input simultaneously to a PDP8 computer and 8-second samples were stored at rest, at the termination of symptom-limited exercise, and in the recovery period. ST isopotential surface maps were subsequently constructed. The presence and praecordial projection of ST abnormality were related to the arteriographic distribution of coronary disease. The ECG data were abnormal in 56 of 58 patients with coronary disease and permitted the identification of left anterior descending artery disease in 49 of 53, right coronary artery disease in 39 of 43, and circumflex artery disease in 24 of 30. Mapping separated those with single vessel from those with multiple vessel disease in 91% of patients with coronary disease. These preliminary results suggest that exercise ECG body surface mapping may provide an attractive non-invasive approach to the investigation of patients with coronary disease.