Aims: To evaluate whether carvedilol influences exercise hyperventilation and the ventilatory response to hypoxia in heart failure (HF). Methods and results: Fifteen HF patients participated to this double blind, randomised, placebo controlled, cross-over study. Patients were evaluated by quality of life questionnaire, echocardiography, pulmonary function and cardiopulmonary exercise tests (ramp and constant workload) both in normoxia (FiO(2) = 21%) and hypoxia (FiO(2) = 16%, equivalent to a simulated altitude of 2000m). Carvedilol improved clinical condition and reduced left ventricle size, but had no effect on lung mechanics. In nonnoxia during exercise, ventilation was lower, V-CO2 unchanged and PaCO2 (constant workload) or PetCO(2) (ramp) higher with carvedilol, exercise capacity was unchanged (peak workload 92 +/- 22 and 90 +/- 22W for placebo and carvedilol, respectively). Abnormal V-E/V-CO2 slope was reduced by carvedilol. Hypoxia increased ventilation but less with carvedilol; exercise capacity decreased to 87 +/- 21 W (placebo) and to 80 +/- 11W (carvedilol,p < 0.01). With hypoxia, carvedilol decreased V-E/V-CO2 slope. At constant workload exercise with hypoxia, PaO2 decreased to 69 +/- 6 mm Hg (placebo) and to 64 +/- 5 (carvedilol,p < 0.01). Conclusion: Carvedilol reduced hyperventilation possibly by reducing peripheral chemoreflex sensitivity as suggested by PaCO2 increase with normoxia and PaO2 decrease with hypoxia without V-CO2, and V-D/V-T changes. Lessening hyperventilation is beneficial when breathing normally, but detrimental when hyperventilation is needed for exercise at high altitude. (c) 2006 European Society of Cardiology. Published by Elsevier B.V. All rights reserved.