Maximum O2 and CO2 fluxes during exercise were less perturbed by hypoxia in Quechua natives from the Andes than in lowlanders. In exploring how this was achieved, we found that, for a given work rate, Quechua highlanders at 4,200 m accumulated substantially less lactate than lowlanders at sea level normoxia (approximately 5-7 vs. 10-14 mM) despite hypobaric hypoxia. This phenomenon, known as the lactate paradox, was entirely refractory to normoxia-hypoxia transitions. In lowlanders, the lactate paradox is an acclimation; however, in Quechuas, the lactate paradox is an expression of metabolic organization that did not deacclimate, at least over the 6-wk period of our study. Thus it was concluded that this metabolic organization is a developmentally or genetically fixed characteristic selected because of the efficiency advantage of aerobic metabolism (high ATP yield per mol of substrate metabolized) compared with anaerobic glycolysis. Measurements of respiratory quotient indicated preferential use of carbohydrate as fuel for muscle work, which is also advantageous in hypoxia because it maximizes the yield of ATP per mol of O2 consumed. Finally, minimizing the cost of muscle work was also reflected in energetic efficiency as classically defined (power output per metabolic power input); this was evident at all work rates but was most pronounced at submaximal work rates (efficiency approximately 1.5 times higher than in lowlander athletes). Because plots of power output vs. metabolic power input did not extrapolate to the origin, it was concluded 1) that exercise in both groups sustained a significant ATP expenditure not convertible to mechanical work but 2) that this expenditure was downregulated in Andean natives by thus far unexplained mechanisms.