Early postexercise polarcardiographic (PCG) changes in the ST segment vector, as reflected in the variable MS X theta, are reported to clearly distinguish normal subjects from those with myocardial ischemia. We prospectively assessed the value of this test in detecting coronary artery disease (CAD) during treadmill exercise in 178 patients within 1 week of diagnostic coronary angiography. The average postexercise MS X theta was 16.9 +/- 9.7 mV degrees in patients with CAD, whereas it was 16.2 +/- 10.2 mV degrees in patients with CAD (p = NS). The optimal ratio of sensitivity to specificity, using different normal-abnormal cutoff values, was 57% and 56%, respectively, for a delineating MS X theta value of 12. The sensitivity and specificity of a simultaneously recorded 14-lead ECG using the criterion of exercise-induced horizontal downsloping ST segment depression greater than or equal to 0.1 mV, was 71% and 78%, respectively. The PCG results were not improved by analyzing MS X theta during exercise, or by analyzing the difference in MS X theta between rest and exercise or rest and postexercise. The sensitivity of the PCG for multivessel or left anterior descending CAD was higher than for less severe forms of CAD, but was significantly less sensitive than the 14-lead exercise ECG (70% versus 84%; p = 0.02; and 62% versus 75%; p = 0.05, respectively). Thus, exercise polarcardiography, as employed in the present study, does not improve the diagnostic content of the 14-lead exercise ECG. Contrary to previous reports, there is no clear separation of normal from abnormal MS X theta values when a prospective series of patients is tested.