Background Chronic heart failure causes disturbances in ventilation and pulmonary gas transfer that participate in limiting peak exercise oxygen uptake (VO_2p ). The β-adrenergic receptor blocker carvedilol improves left ventricular (LV) function and not VO_2p . This study was aimed at investigating the pulmonary response to changes in LV performance produced by carvedilol in patients with chronic heart failure. Methods Twenty-one patients with New York Heart Association class II to III heart failure were randomly assigned (2 to 1) to carvedilol (25 mg twice daily, n = 14) or placebo (n = 7) for 6 months. Rest forced expiratory volume (FEV₁ ), vital capacity, total lung capacity, carbon monoxide diffusing capacity, its alveolar-capillary membrane component, pulmonary venous and transmitral flows (for monitoring changes in LV end-diastolic pressure), LV diastolic and systolic dimensions, stroke volume, ejection fraction, and fiber shortening velocity were measured at baseline and at 3 and 6 months. VO_2p , peak ratio of dead space to tidal volume (VD/VT_p ), ventilatory equivalent for carbon dioxide production (VE/VCO₂ ), and VO₂ at anaerobic threshold (VO_2at ) were also determined. Results FEV₁ , vital capacity, total lung capacity, carbon monoxide diffusing capacity, and the alveolar-capillary membrane component were impaired in chronic heart failure compared with 14 volunteers and did not vary with treatment. Carvedilol reduced end-diastolic pressure, end-diastolic diameter, and end-systolic diameter and increased ejection fraction, stroke volume, and fiber shortening velocity without affecting VO_2p , VO_2at , VD/VT_p , or VE/VCO₂ at 3 and 6 months. Placebo did not produce significant changes. Conclusions In chronic heart failure carvedilol ameliorates LV function at rest and does not significantly affect ventilation and pulmonary gas transfer or functional capacity. These results suggest that improvement in cardiac hemodynamics with carvedilol does not reverse pulmonary dysfunction. Persistent lung impairment might have some role in the failure of carvedilol to improve exercise performance. (Am Heart J 1999;138:460-7.)