Objective: To determine by 31P nuclear magnetic resonance (NMR) spectroscopy the efficacy of training in improving aerobic metabolism of calf muscle in nonamputated limb after recent vascular amputation; to assess the possible associated microcirculatory changes; and to evaluate the need for noninvasive monitoring techniques during training in the nonamputated limb after recent vascular amputation. Design: Prospective study, before and after training. Subjects served as their own controls and were compared with a control group. Setting: Rehabilitation center of a university hospital. Patients: Ten unilateral vascular amputated patients were included with ankle systolic index between 0.5 and 0.8 in the nonamputated limb, and 10 control subjects without cardiovascular disease or risk factors of atherosclerosis with ankle systolic index of >.95. Intervention: Walking with prosthesis at self-selected velocity over increasing walking distance, arm training at a workload of 60% of a maximal arm test, and analytical exercises of the nonamputated leg (dynamic contractions against low resistance). Subjects received training as inpatients, 5 days a week. Main Outcome Measures: Before and after training, ankle systolic index, forefoot transcutaneous oxygen tension (TcPo 2) and veno-arteriolar reflex, and digital plethysmography of the second toe with reactive hyperemia test were studied. Changes in calf muscle pH, phosphocreatine (PCr), and inorganic phosphate (Pi) were measured by 31P NMR spectroscopy at rest and during a plantar flexion-type incremental protocol. Results: There was no significant difference in ankle systolic index (.63 ± .10 vs .64 ± .07) or in TcPo 2 (42 ± 11 vs 44 ± 10mmHg), and there was reappearance of veno-arteriolar reflex in 3 cases, of a plethysmographic signal in 2 cases, and of the positivity of the reactive hyperemia test in 3 cases. No differences were found with 31P NMR spectroscopy at rest before and after training. At the same workload (1 watt) the difference of the ratio (PCr/(PCr + Pi)) of rest to effort (PCr depletion) was significantly increased in the amputated patients (.423 ± .159 vs .145 ± .058; p < .01). This difference of ratio was lower after training (.360 ± .158 vs .423 ± .159; p < .05). The pH was less acid between the two periods. Conclusion: Vascular monitoring with systolic index and TcPo 2 is necessary to follow and to prevent serious ischemia of the nonamputated limb. Claudication is often not detected because of early exhaustion during walking. Training after recent vascular amputation improves the skeletal muscle oxydative capacity.