To assess the prevalence of mental stress-induced myocardial ischemia and investigate the pathogenetic mechanisms by which emotional stress may induce myocardial ischemia, we studied 372 patients with angina pectoris who underwent mental arithmetic and exercise stress testings. Hyperventilation tests were also performed in 176 patients, and 340 patients underwent coronary arteriography. Sixty-one patients showed significant ST segment abnormalities during mental arithmetic and exercise stress testings (group 1). Two hundred eleven patients had negative responses to mental stress but positive exercise tests (group 2), whereas both tests were negative in 100 patients (group 3). Mental stress induced significant increases in heart rate and systolic blood pressure in the three groups of patients; however, group 1 patients had higher increases in rate-pressure product (mm Hg x beats/min) than group 2 and group 3 patients (14,909 +/- 3,894 versus 12,985 +/- 2,900 versus 12,724 +/- 4,400 mm Hg x beats/min, p < 0.01). Group 1 patients had shorter exercise durations than group 2 or group 3 (4.06 +/- 1.55 versus 7.65 +/- 3.07 versus 13.9 +/- 5.31 minutes, p < 0.01), although rate-pressure products at peak exercise were similar in groups 1 and 2 (20,277 +/- 6,058 versus 20,768 +/- 3,864, p = NS) and significantly higher in group 3 (26,221 +/- 7,100/mm Hg x beats/min, p < 0.01). In group 1 patients, rate-pressure product at the ischemic threshold during physical exercise did not differ from rate-pressure product attained during positive mental stress (14,742 +/- 1,816 versus 14,909 +/- 3,894 mm Hg x beats/min, p = NS), whereas rate-pressure product in group 2 patients achieved at the ischemic threshold during exercise was significantly higher than that achieved during negative mental stress (17,221 +/- 2,450 versus 12,985 +/- 2,900 mm Hg x beats/min, p < 0.01). In patients with early positive responses to hyperventilation and positive mental stress tests, rate-pressure products at the ischemic threshold during hyperventilation, mental stress, and exercise were similar (15,121 +/- 5,222 versus 15,693 +/- 2,499 versus 16,624 +/- 3,398 mm Hg x beats/min, p = NS), whereas in the two patients with delayed positive hyperventilation and positive mental stress tests, rate-pressure products of the ischemic threshold during hyperventilation and mental stress were similar but lower than that at ischemic threshold during exercise. Our findings demonstrate that mental stress induces electrocardiographic signs of myocardial ischemia in 18% of patients with coronary artery disease. The sharp increases in heart rate and blood pressure in patients showing positive responses to mental and exercise stress suggest a hyperreactive response to both stress tests. Increased vasomotor tone may also play a role in modulating the ischemic response to mental stress limiting coronary vasodilation when demand-related ischemia occurs, although a clear vasospasm appears to be a rare phenomenon.