To assess the value and predictive limitations of the exercise ST/HR slope, exercise test results were compared in 50 patients with stable angina and in 17 normal subjects with those in two groups known to have a high prevalence of inaccurate electrocardiographic responses to exercise. The last two groups included 51 patients tested within three weeks of acute myocardial infarction and 17 with important aortic regurgitation but no coronary disease. Of the normal subjects, 16 (94%) had ST/HR values less than or equal to 1 X 1 microV/beat/min. Of those with stable angina pectoris, 42 of 46 (91%) patients with coronary artery disease had ST/HR slopes ranging from 1 X 2 to 20 X 0 microV/beat/min, with false negative findings (slopes less than or equal to 1 X 1 microV/beat/min) in only four (9%). In contrast, of those with recent myocardial infarction, 15 of 42 (36%) with coronary disease had false negative slopes, including 12 of 20 (60%) with anterior wall injury. Of those with aortic regurgitation, conversely, 14 of 16 (88%) patients with calculable ST/HR slopes had values greater than 1 X 1 microV/beat/min despite the absence of coronary disease. Despite the accuracy of the test in patients with stable angina, false negative results are common in those after recent myocardial infarction, and false positive results occur often in those with abnormal volume loading due to aortic regurgitation.