In two groups of patients the detection of myocardial infarction (MI) by analysis of four different electrocardiographic methods was evaluated. The various methods included the conventional 12 lead ECG (CV-ECG), the 12 lead ECG derived from Frank XYZ lead system signals (D-ECG), the polarcardiogram (PCG) and the vectorcardiogram (VCG). An invasive group consisted of 137 patients who had undergone cardiac catheterization. An MI was defined as a regional wall motion abnormality in the distribution of a coronary artery with at least 70% diameter reduction. The noninvasive group consisted of 116 patients in whom independent clinical information was limited to noninvasive assessments. In this group, Telemed Computer Systems' interpretation of the conventional (TC-ECG) and derived (TD-ECG) electrocardiogram was also available for comparison. An MI was defined in this group as either a compatible history with documented cardiac enzyme elevations, a resting defect on thallium scan, or a regional wall motion abnormality in a resting, radionuclide isotope ventriculogram. In this study the other methods of ECG evaluation demonstrated no advantage over the electrocardiographer's reading of the conventional ECG.