A randomized double-blind study using the double-dummy procedure was carried out to compare the anti-ischemic and antianginal effects of bupranolol delivered transdermally (from patches each containing 30 mg) and propranolol administered orally (160-mg slow-release capsule) with respect to their anti-ischemic and antianginal effects. Forty patients, 23 men and 17 women, with coronary heart disease demonstrated by coronary angiography and stable angina pectoris took part in the study. They were assigned at random to two groups of equal size. At the time of the study, coronary surgery was impossible or unnecessary in these patients. After a 1-week wash-out phase, the patients were treated with a bupranolol transdermal therapeutic system, 30 mg/d (group I), or with slow-release propranolol, 160 mg/d (group II), for 8 weeks. To ensure that the study was truly double-blind, the patients in group II also took one placebo capsule each day, and the patients in group I applied a placebo patch daily. On the last day of the wash-out phase and after 4 and 8 weeks' treatment, the patients underwent standardized exercise tests on an ergometer up to the point at which the test had to be stopped. A 24-hour ECG was recorded, and laboratory values were measured. The frequency of anginal attacks and the patients' consumption of glycerol trinitrate to relieve angina were recorded throughout the study. At a comparable workload the ST-segment depression was reduced on average by 61.5% in group I (P < 0.001) and by 37% in group II (P < 0.001). The difference between the groups was due to more than just chance (P < 0.025). The frequency of anginal attacks decreased, on average by 75% in group I (P < 0.001) and by 37.9% in group II (P < 0.001). The difference between the two groups was statistically significant (P < 0.001). The consumption of glycerol trinitrate to relieve angina fell, on average, by 79.1% in group I (P < 0.001) and by 36% in group II (P < 0.001). The difference between the two groups was statistically significant (P < 0.001). The number of ischemic episodes (> 0.1 mV > 1 min) in the Holter-ECG fell, on average, by 70.9% in group I (P < 0.001) and by 34.1% in group II (P < 0.001). The difference between the groups was statistically significant (P < 0.01).