The extent and time-course of changes in lung volumes, ventilatory efficiency at rest and during exercise, and respiratory muscle function and their influence on exercise limitation in congestive heart failure (CHF) are unclear. It is unknown whether respiratory muscle function may predict changes in exercise limitation or may be impaired in patients with poor outcome. 145 male patients (54 +/- 1 years) suffering from CHF (NYHA class I-III, mean 2.3 +/- 0.1), with a LVEF of 23 +/- 1%, and a mean peak O2 uptake (VO2peak) 15.0 +/- 0.5 mL X min(-1) X kg(-1), were studied. They were grouped in Weber functional classes A to D according to their VO2peak. Significant increases in ventilatory equivalents for O2 and CO2 (VE/VCO2peak) and in dead space ventilation at rest and during exercise were found with increasing exercise limitation. Moreover, there was a correlation of static and dynamic lung volumes (inspiratory vital capacity, IVC, r = 0.43, P < 0.01), as well as of maximal inspiratory pressure (MIP; r = 0.46, P < 0.01) with VO2peak. Patients who died (n = 26) or were heart transplanted (n = 20) during a follow-up (mean 800 +/- 10 days) had a reduced MIP (6.4 +/- 0.4 kPa) as compared with survivors (n = 82; 9.3 +/- 0.7 kPa, P < 0.01). In a subgroup of 33 patients re-evaluated after six months, individual changes in IVC and VE/CO2peak, but not in MIP, correlated to changes in VO2peak (r = 0.69 and r = 0.72 respectively; P < 0.01). In CHF, exercise limitation is associated with reversible lung restriction and inefficient ventilation at rest and during exercise. Patients with severe CHF have a significant reduction in MIP, a finding that is associated with poor outcome.