BACKGROUND There is a relationship between obstructive sleep apnea (OSA) and heart failure (HF). Peak oxygen consumption (peak VO(2)), heart rate recovery, and N-terminal-pro-BNP (NT-proBNP) are strong prognostic predictors in HF. The effects of nasal continuous positive airway pressure (nCPAP) on these parameters in OSA patients are not well defined. METHODS Forty patients with newly diagnosed OSA [apnea-hypopnea index (AHI) 37 (20-65) h(-1)] underwent cardiopulmonary exercise testing for assessment of peak VO(2) and heart rate recovery at one (HRR-1) and two (HRR-2) minutes after exercise termination as well as NT-proBNP measurement at baseline and after 7.9+/-1.4 months of effective nCPAP (nightly usage>3.5 h). The effects of nCPAP were compared in patients with mild-to-moderate (AHI<30 h(-1); n=16) vs. severe (AHI>or=30 h(-1); n=24) OSA. RESULTS In the group as a whole, peak VO(2) (baseline: 31.9+/-9.3 vs. follow-up: 33.7+/-9.0 ml/kg/min; p=0.02) and HRR-2 [38 (32-43) vs. 42 (32-47) bpm; p=0.01] but not HRR-1 [22 (15-26) vs. 22 (16-27) bpm; p=0.16] improved from baseline to follow-up. The effect on peak VO(2) was mainly driven by a trend towards an increase in patients with mild-to-moderate OSA (31.8+/-10.7 vs. 33.9+/-10.2 ml/kg/min; p=0.08), whereas an effect on HRR-1 [20 (15-23) vs. 21 (16-26) bpm; p=0.03] and HRR-2 [38 (29-42) vs. 42 (33-47) bpm; p=0.004] was observed only in those with severe OSA. NT-proBNP levels remained unchanged [21 (11-45) vs. 26 (5-52) pg/ml; p=0.6]. CONCLUSIONS Treatment with nCPAP is associated with an improvement in peak VO(2) and heart rate recovery in patients with OSA.