Normal values for heart rate-adjusted indexes of ST segment depression during treadmill exercise electrocardiography (the ST segment/heart rate slope and the delta ST segment/heart rate index) were derived from evaluation of 150 subjects with a low likelihood of coronary artery disease, including 100 normal subjects and 50 subjects with nonanginal chest pain. Partitions chosen by the method of percentile estimation to include 95% of normal subjects remained highly specific in subjects with nonanginal pain syndromes. Sensitivities of the derived partitions for detection of myocardial ischemia were tested in an additional 150 patients with a high likelihood of coronary disease, including 100 patients with angiographically demonstrated coronary obstruction and 50 patients with stable angina. In contrast to the 68% (102 of 150 subjects) sensitivity of standard exercise electrocardiographic criteria for the detection of disease in this population, the sensitivity of an ST segment/heart rate slope partition of 2.4 muV/beats/min was 95% (142 of 150 subjects, p less than 0.001), and the sensitivity of a delta ST segment/heart rate index partition of 1.6 muV/beats/min was 91% (137 of 150 subjects, p less than 0.001). Analysis of receiver-operating curves confirmed the superior performance of the heart rate-adjusted indexes throughout a wide range of test specificities. These findings suggest that heart rate adjustment of ST segment depression can markedly improve the clinical usefulness of the treadmill exercise electrocardiogram.