Of all tissues of the extremities, muscle is the least tolerant of ischemia. Hypothermia of tissue is considered beneficial for the maintenance of viability of muscle in amputated limbs before surgical replantation, but it has never been established that conventional cooling in an ice bath or its equivalent (temperature of tissue, approximately 1 degree Celsius) is the optimum level of hypothermia for minimizing metabolic derangement in ischemic muscle. In this study, we first defined the time course and level of metabolic derangement of muscle in twenty-eight ischemic hind limbs in cats at 22, 15, 10, 5, and 1 degree Celsius. The levels of adenosine triphosphate and phosphocreatine and the mean intracellular pH of the muscles in the lateral aspect of the thigh in each limb were monitored with phosphorus nuclear magnetic-resonance spectroscopy over time. The excised muscles from six freshly amputated legs of live humans were then similarly studied to determine whether muscles from cats and from humans exhibit comparable bioenergetic responses to hypothermic ischemia. A final series of ten ischemic hind limbs from cats was studied by nuclear magnetic resonance and muscle biopsy for direct biochemical assay of tissue energy metabolites to compare the metabolic benefits of two different methods of preserving limbs: continuous cooling in an ice bath, and a newly devised protocol for the rapid induction and maintenance of so-called intermediate (10 +/- 5 degrees Celsius) hypothermia of tissue. Ischemic skeletal muscle in cats exhibited a paradoxical metabolic response to extreme cold (1 degree Celsius). The rate of metabolic deterioration progressively declined with decreasing temperature of tissue to 10 degrees Celsius. However, at 5 degrees Celsius, no additional benefit was detected, and at 1 degree Celsius, there was a significant acceleration in the rates of degradation of adenosine triphosphate and phosphocreatine and in the production of lactate. The rate of degradation of adenosine triphosphate in human ischemic muscle was also faster at 1 degree Celsius than at 10 degrees Celsius. This paradoxical response is apparently due to a severe inhibition of the calcium pump of the sarcoplasmic reticulum of the muscle cell at temperatures of less than 5 degrees Celsius. The inhibition permits an efflux of calcium to the myofibrils, which stimulates both glycolysis and the degradation of adenosine triphosphate by myofibrillar adenosine triphosphatase.