The purpose of this study was to delineate among the usually gathered parameters in an electrocardiographic exercise test the determinants of its positive outcome (delta ST decreases > or = 1 mm measured at 80 msec from the J point). The authors studied 832 patients investigated with Bruce's exercise testing and with diagnostic coronary arteriography, all of whom were shown to have significant coronary artery disease (diameter stenosis > or = 50%). Twenty-five demographic, clinical, electrocardiographic, exercise, and anatomic/hemodynamic parameters were analyzed. The stepwise forward logistic regression analysis retained seven among them as significant independent predictors: four as positive contributors: (1) three-vessel and/or left main disease (P = 0.0000), (2) Gensini's angio-graphic score for disease extent (P = 0.0025), (3) anginal pain during the test (P = 0.0000), and (4) age (P = 0.0031) and three as negative contributors: (1) resting heart rate (P = 0.0004), (2) history of old myocardial infarction (P = 0.0019), and (3) pathological Q waves at the resting ECG (P = 0.0018). These findings establish standards that permit the prediction of the positive electrocardiographic exercise outcome in patients with significant coronary disease.