A new stress criterion for diagnosing coronary artery disease (CAD) is proposed. This criterion is based upon the measurement of the ratio of systolic blood pressure to heart rate (SBP/HR) in the postexercise period. Bicycle exercise stress tests on 186 patients and coronary angiograms of 173 of them were studied. Three minutes after exercise, patients with CAD, without myocardial infarction showed significantly higher ratios than patients without CAD (1.72 +/- 0.48 vs. 1.28 +/- 0.25; p < 0.01). Failure of the ratio to decrease below 1.38 3 min after the exercise was considered an abnormal response. The sensitivity, specificity and predictive accuracy of this criterion were 82, 81, 81 %, respectively; the corresponding values for S-T depression were 79, 48 (p < 0.01 versus SBP/HR), 67% (p < 0.01 vs. SBP/HR). By this new criterion, it was possible to identify 72% of false-positive S-T segment responses to stress. Combining an abnormal ratio with a positive S-T segment response failed to detect only 2 % of patients with multivessel disease. Therefore, SBP/HR represents an additional and valuable criterion for diagnosing CAD by exercise with or without associated electrocardiographic changes. Finding a normal ratio in patients with positive S-T segment depression should alert to the possibility of a false-positive response. The finding of a normal ratio and a normal S-T response allows to rule out with very high probability the presence of multivessel disease.