Patients with heart failure frequently report leg fatigue during exercise. At present, however, there is no objective method of detecting leg muscle abnormalities in such patients. To determine if phosphorus-31 nuclear magnetic resonance spectroscopy can provide such information, this technique was used to compare calf responses to stair climbing and plantarflexion in 20 patients with heart failure (peak oxygen consumption (VO2) of 13.6 +/- 5 ml/kg/min, ejection fraction 20 +/- 5%) and 9 age-matched normal subjects. Work was quantified by measuring VO2. At rest, both groups exhibited similar inorganic phosphorus to phosphocreatine (Pi/PCr) ratios (patients with heart failure 0.21 +/- 0.07, normal subjects 0.21 +/- 0.06, difference not significant) and pH levels (patients with heart failure 7.06 +/- 0.17, normal subjects 7.05 +/- 0.11, difference not significant). In both normal subjects and patients with heart failure, exercise resulted in a progressive rise in Pi/PCr as VO2 increased. However, examination of the relation of VO2 versus Pi/PCr revealed steeper slopes in patients with heart failure during both stair climbing and plantar-flexion. Neither form of exercise decreased calf pH in normal subjects. In the patients with heart failure, significant decreases in pH were noted during the highest work level of plantarflexion (pH of heart failure patients 6.86 +/- 0.20, pH of normal subjects 7.07 +/- 0.14, p less than 0.01). Metabolic recovery time was also prolonged in the patients with heart failure versus normal subjects (3.3 +/- 0.8 vs 2.1 +/- 0.5 minutes, respectively, p less than 0.002). These findings indicate that phosphorus-31 nuclear magnetic resonance provides objective evidence of leg muscle abnormalities in patients with heart failure.(ABSTRACT TRUNCATED AT 250 WORDS)