Episodes of transient myocardial ischemia during ambulatory activities are common in patients with stable coronary artery disease and who are often asymptomatic. Selection of therapy for episodes of asymptomatic ischemia is limited by a lack of direct comparative studies. To determine the most effective monotherapy for patients with stable angina and a high frequency of asymptomatic ischemic episodes, propranolol-LA (mean daily dose, 293 mg), diltiazem-SR (mean daily dose, 350 mg), nifedipine (mean daily dose, 79 mg) were each compared with placebo, each for 2 weeks, in a randomized, double-blinded, crossover trial. Entry criteria were a positive exercise treadmill test during placebo therapy characterized by 1.0 mm or more ST segment depression and angina pectoris, and six or more episodes of transient ST segment depression of 1.0 mm or more on a 48-hour ambulatory electrocardiogram. One hundred ninety-four patients were screened, 63 were eligible and received randomized therapy, of which 56 patients completed at least two of the four treatment periods and were included in an intent-to-treat analysis. Fifty patients completed all four treatment phases and were included in the protocol-completed analysis. Anti-ischemia efficacy was assessed by 48-hour ambulatory electrocardiographic monitoring, exercise treadmill tests, and anginal diaries. Ninety-four percent of all episodes of ambulatory ischemia were asymptomatic. Compared with placebo, only propranolol was associated with a marked reduction in all manifestations of asymptomatic ischemia during ambulatory electrocardiographic monitoring (2.3 versus 1.0 episodes/24 hr; mean duration of ischemia per 24 hours, 43.6 versus 5.7 minutes; both p < 0.0001). Diltiazem's reduction of the frequency of episodes compared with placebo (2.3 versus 1.9 episodes/24 hr) was associated with a trend (p = 0.08) in the protocol-completed analysis and with a significant reduction in the intent-to-treat analysis (p = 0.03). Nifedipine had no significant effects on any measured variable of ambulatory ischemia. The dosages of medication used may have been excessive for some patients, and a more beneficial effect may have been evident at a lower dose. In contrast to the marked effects of the active agents on ambulatory asymptomatic ischemia, the effects on exercise performance and angina pectoris were slight. The active agents modestly improved treadmill exercise duration time until 1 mm ST segment depression (3%), and only propranolol and diltiazem had significant effects. Only diltiazem significantly prolonged the total exercise time. Anginal frequency was significantly decreased by both propranolol and diltiazem. We conclude that propranolol is effective for treating episodes of asymptomatic ischemia in patients with stable angina and that the magnitude of improvement in ambulatory asymptomatic ischemia is far greater than the improvement in exercise performance and angina symptoms.