This study evaluated the Moxus metabolic system with the Douglas bag method (DBM) as criterion. Reliability and validity were investigated in a wide range of ventilation and oxygen uptake and two sensors for determining ventilation were included. Thirteen well-trained athletes participated in one pre-test and four tests for data collection, exercising on a cycle ergometer at five submaximal powers (50-263 W) and at VO₂max. Gas exchange variables were measured simultaneously using a serial setup with data collected on different days in an order randomized between Moxus with pneumotachometer (MP) and turbine flowmeter (MT) sensors for ventilation. Reliability with both sensors was comparable to the DBM. Average CV (%) of all exercise intensities were with MP: 3.0 ± 1.3 for VO₂, 3.8 ± 1.5 for VCO₂, 3.1 ± 1.2 for the respiratory exchange ratio (RER) and 4.2 ± 0.8 for V E. The corresponding values with MT were: 2.7 ± 0.3 for VO₂, 4.7 ± 0.4 for VCO₂, 3.3 ± 0.9 for RER and 4.8 ± 1.4 for V E. Validity was acceptable except for small differences related to the determination of ventilation. The relative differences in relation to DBM at the powers including VO₂max were similar for both sensors with the ranges being: +4 to -2 % for V E, +5 to -3 % for VO₂ and +5 to -4 % for VCO₂ while RER did not differ at any power. The Moxus metabolic system shows high and adequate reliability and reasonable validity over a wide measurement range. At a few exercise levels, V E differed slightly from DBM, resulting in concomitant changes in VO₂ and VCO₂.