To examine the effects of propranolol and nifedipine on exercise-induced attack in patients with variant angina, exercise 20`T1 myocardial scintigraphy with quantitative analysis by emission-computed tomography was performed in 20 patients with variant angina after oral propranolol (80 mg), nifedipine (20 mg), and placebo. Exercise-induced attack occurred in 11 patients on placebo, in 14 on propranolol, and in none on nifedipine. The exercise duration was significantly shorter in those on propranolol (p < .05), but significantly longer in patients on nifedipine (p < .05) than in those on placebo. The peak rate-pressure product was significantly lower in patients on propranolol (p < .01), but did not change in those on nifedipine, as compared with that in patients on placebo. The size of the perfusion defect as measured by 201TI tomography was significantly greater in patients on propranolol (p < .05), but significantly less in those on nifedipine (p < .01) than in those on placebo. In conclusion, propranolol does not suppress but rather may aggravate exercise-induced attack in patients with variant angina, while nifedipine suppresses it. This unfavorable effect of propranolol on exercise-induced attack in patients with variant angina is likely to be due to a reduction of regional myocardial blood flow.