To compare the effects of two beta-blocking drugs: a nonselective (propranolol) and a cardioselective with modest intrinsic sympathomimetic activity (visacor), 24 patients with stable angina pectoris performed a control exercise (without medication) on a bicycle ergometer (increments of 30 W every 3 min), and thereafter were randomized to receive either propranolol (40 mg t.i.d.) or visacor (200 mg once daily) for a 48-hour double-blind trial. The 2 groups on control exercise were similar with regard to their exercise tolerance: 7.6 ± 2.3 versus 7.1 ± 1.4 min (NS) and the behavior of heart rate, systolic, diastolic blood pressure and double product, at rest and during exercise. They exercised on the 2nd day 2 h after the intake of propranolol or visacor. In the 12 patients randomized to propranolol, heart rate, systolic and diastolic pressures, double product were significantly reduced at rest, compared with control exercise: 67 ± 8 versus 81 ± 10 beats/min (p < 0.01), 132 ± 20 versus 146 ± 21 mm Hg (p < 0.02), 80 ± 8 versus 88 ± 10 mm Hg (p < 0.02), 8,828 ± 1,927 versus 11,863 ± 2,138 mm Hg·min-1 (p < 0.001), respectively. On the contrary, in the 12 patients randomized to visacor, these parameters at rest were less modified and only heart rate was significantly decreased: 71 ± 9 versus 81 ± 11 beats/min (p < 0.05). The onset of ST-segment depression (≧ 1.0 mm) was delayed 14% (p < 0.05) in the propranolol group, and 31 % (p < 0.01) in the visacor group. The peak double product decreased 27% (p < 0.001) in the propranolol group and 17% (p < 0.02) in the visacor group. Both drugs increased exercise duration, but only in the visacor group, the changes were significant (17%, p < 0.02, vs. 5%, NS in propranolol group). In conclusion, visacor seems to be at least as efficient as propranolol in exercise-induced angina.