Background: beta-blockade controls the ventricular response to exercise in chronic atrial fibrillation (AF), but the effects of beta-blockers on exercise capacity in AF have been debated. Methods: Twelve men with AF (65 +/- 8 years) participated in a randomized, double-blind, placebo-controlled study of betaxolol (20 mg daily). Patients underwent maximal exercise testing with ventilatory gas exchange analysis, and a separate, submaximal test (50% of maximum) during which cardiac output was measured by a CO2 rebreathing technique. Results: After betaxolol therapy, heart rate was reduced both at rest (92 +/- 27 vs 62 +/- 12 beats/min; p < 0.001) and at peak exercise (173 +/- 22 vs 116 +/- 24 beats/min; p < 0.001). Maximal oxygen uptake ((V) over dot O-2) was reduced by 19% after betaxolol (21.8 +/- 5.3 with placebo vs 17.6 +/- 5.1 mL/kg/min with betaxolol; p < 0.05), with similar reductions observed for maximal exercise time, minute ventilation, and CO2 production. (V) over dot O-2 was reduced by a similar extent (19%) at the ventilatory threshold. Submaximal cardiac output was reduced by 15% during betaxolol therapy (12.9 +/- 2.3 vs 10.9 +/- 1.3 L/min; p < 0.05), and stroke volume was higher (88.0 +/- 21 vs 105.6 +/- 19 mL/beat; p < 0.05). Conclusion: Betaxolol therapy in patients with AF effectively controlled the ventricular rate at rest and during exercise, but also caused considerable reductions in maximal (V) over dot O-2 and cardiac output during exercise. The observed increase in stroke volume could not adequately compensate for reduced heart rate to maintain ire, during exercise.