A new system for continuous measurement of oxygen uptake was tested. The tidal volumes delivered by the system, the accuracy of the oxygen sensor cell, measurement of the oxygen uptake at different inspired oxygen concentrations, and comparative tests of oxygen uptake using the Fick principle were studied. Judging from our results, even thorough maintenance could not entirely prevent leakages. Due to a construction error, there was a large increase in measured oxygen uptake if inspired oxygen concentration was increased. While carbon‐dioxide output was constant, this led to respiratory quotients of less than 0.5. By modifying the system with a mixing chamber for the inspiratory gas, new tubing to the oxygen sensor cell and changes in the computer software, these errors have been corrected in the meantime. In comparative tests of oxygen uptake using the Fick principle (cardiac output and arteriovenous oxygen content difference), we found a mean difference between EMC and the Fick principle of about 6.6% (P≤0.001, r = 0.91) with nearly identical standard deviations for both values. Our experience indicates that the EMC is suitable for continuous measurements of oxygen uptake. Additional technical improvements seem to be possi