Patients with advanced heart failure have bioenergetic abnormalities of skeletal muscle metabolism during exercise. Using 31P magnetic resonance spectroscopy, we sought to determine whether skeletal metabolic responses to exercise are normalized by orthotopic cardiac transplantation. Four groups were studied: healthy normal volunteers (n = 9), subjects awaiting heart transplantation (n = 10), subjects < 6 months (mean, 4 months) after transplant (n = 9), and subjects > 6 months (mean, 15 months) after transplant (n = 8). None of the posttransplant patients had biopsy evidence of rejection at the time of study. There were no significant differences in age, preoperative functional class, or symptom duration among the three patient groups. Metabolic responses were monitored in the dominant arm during incremental weight pull exercise and 10 minutes of recovery by 31P magnetic resonance spectroscopy, with measurement of pH and the phosphocreatine (PCr)/(PCr + inorganic phosphate [Pi]) ratio, an index of PCr concentration. In addition, based on recovery data, the rate of PCr resynthesis was calculated as a measure of oxidative metabolism that is independent of work level, recruitment, or muscle mass, and the effective maximal rate of mitochondrial ATP synthesis (Vmax) was determined. Analysis was by ANOVA. There were no differences between groups in pH or PCr/(PCr + Pi) at rest. Compared with the normal control group, the pretransplant group had a decreased exercise duration (11.3 +/- 2.5 versus 15.0 +/- 1.3 minutes, P = .02), a lower submaximal exercise PCr/(PCr + Pi) ratio (0.58 +/- 0.11 versus 0.76 +/- 0.08, P < .05), a reduced PCr resynthesis rate (13 +/- 6 versus 22 +/- 9 mmol/L per minute, P < .05), and a lower calculated Vmax (26 +/- 14 versus 53 +/- 26 mmol/L per minute, P < .05). In the group studied early after transplantation, all the changes noted in the pretransplant group persisted and were if anything somewhat worse. In the group studied late after transplantation, there was a significant improvement in the PCr resynthesis rate compared with the early-posttransplant group (27 +/- 6 late versus 15 +/- 6 mmol/L per minute early, P < .05) and statistically nonsignificant trends toward improvements in submaximal exercise pH (6.86 +/- 0.24 late versus 6.72 +/- 0.24 early) and submaximal PCr/(PCr + Pi) ratio (0.56 +/- 0.14 late versus 0.44 +/- 0.15 early) and Vmax (45 +/- 21 late versus 33 +/- 15 mmol/L per minute early). However, compared with normal subjects, exercise duration and submaximal PCr/(PCr + Pi) were still reduced in the late-posttransplant group. Despite successful heart transplantation, skeletal muscle abnormalities of advanced heart failure persist for indefinite periods, although partial improvement occurred at late times. The persistent abnormalities may contribute to the reduced exercise capacity that is present in most patients after transplantation.