To analyze the usefulness of a single exercise test to predict the presence of fixed obstructive coronary artery disease in patients with active coronary spasm, 91 consecutive patients with angiographically proven symptomatic coronary artery spasm who had performed a symptom-limited exercise test within the week before diagnostic coronary angiography were studied. Coronary angiography revealed significant coronary obstructions in 61 patients (67%). According to the type of angina, the prevalence of significant coronary stenosis was 53% for patients with angina at rest, 68% for those with effort angina, and 92% for those with mixed angina. Exercise-induced ST segment elevation was present in eight patients (9%), ST segment depression was seen in 37 patients (41%), and no ST abnormalities in 46 (50%). There was not a significant relationship between the ST segment response to exercise and the clinical variables assessed except for coronary anatomy. Abnormal exercise test results were significantly more frequent in patients with significant coronary obstructions than in those without significant coronary occlusions (62% versus 23%; p less than 0.01). ST elevation was not useful to predict the presence of fixed coronary lesions. However, ST depression strongly suggested the presence of underlying coronary lesions with a sensitivity of 54%, a specificity of 87%, and a positive predictive value of 89%. Using this criterion, 65% of the patients were correctly classified. The results indicate that despite the functional component of ischemia in patients with coronary spasm, ST segment depression with exercise is still a highly specific sign with a high positive predictive value for the presence of significant coronary artery disease.(ABSTRACT TRUNCATED AT 250 WORDS)