The aim of this study was to test the linearity, precision, and accuracy of the measurements made by the K2 system at sea level (SL) and moderate altitude (MA) (barometric pressure = 591.5 +/- 0.5 mm Hg). To minimize the day-to-day biovariability, a testing protocol based on repeated-alternated measures was used at rest and during three levels of submaximal exercice lasting 12 min each, at 25%, 50%, 75% of the peak workload. The measurements of the respiratory parameters were compared with those obtained with a metabolic measurement cart. The results reported in this study show that the K2 system was an accurate and consistent system for oxygen uptake (VO2) measurements at SL. The K2 system was consistent at MA; however, the K2 system significantly overestimated and underestimated the VO2 computations at rest and 25% of the peak workload, respectively. The calculation of VO2 using the K2 system which assumes that RER = 1.00 had specific effects for the calculation of oxygen uptake. The measurements of FEO2 selectively differed from those obtained with the metabolic measurement cart at MA. Therefore, we concluded that the K2 system was an accurate system for VO2 measurements during submaximal exercices (50%-75% of the peak workload) under laboratory conditions at MA (up to 2,000 m).