Several studies of phosphorus 31 (P-31) magnetic resonance spectroscopy (MRS) have demonstrated the presence of skeletal muscle metabolic abnormalities during exercise in patients with chronic heart failure (CHF). We studied the contribution of these abnormalities to the limitation of exercise capacity in CHF. ln 25 patients (age 57 +/- 2 years, left ventricular ejection fraction [LVEF] 28% +/- 1.6%, peak oxygen consumption (Vo(2)) 16 +/- 1.2 ml/kg/mm) (mean +/- SEM), we studied the calf muscle at rest and during plantar flexion with P-31 MRS. The phosphocreatine (PCr) depletion rate was significantly negatively correlated to peak Vo(2) (r = -0.62, p = 0.001) but not to LVEF. Muscle pH was correlated with the inorganic phosphorus (Pi)/PCr ratio (r = -0.69, p = 0.0001) and with the PCr/adenosine triphosphate beta (ATP beta) ratio (which negatively relates to adenosine diphosphate [ADP] concentration) (r = 0.65, p = 0.00001). Although muscle ATP (ATP/sum of phosphorus [Sigma P] remained stable, in 8 patients ATP/Sigma P decreased significantly (-15% +/- 4%, p = 0.0002). In this ATP-depleted group, peak ire, was significantly lower than that of the nondepleted group and PCr depletion more rapid, whereas LVEF did not differ. Skeletal muscle metabolic abnormalities in CHF contribute markedly to the alteration of exercise capacity. Rapid PCr depletion and muscle acidosis are the most relevant abnormalities. ATP depletion and excessive increase in ADP during exercise may contribute further to exercise limitation specifically in patients with more marked CHF.