This study evaluated an ergo-spirometry system based on mixed expired gas for gas analyses and an inspiratory based determination of flow. There were 74 paired samples of oxygen uptake (VO(2)) and related variables including pulmonary ventilation (V(E)), fractional concentrations of expired CO(2) and O(2) ( F(E)CO(2) and F(E)O(2), respectively), as well as CO(2) output (VCO(2)) which were obtained from the metabolic cart and a Douglas bag system during 22 min submaximal and 5-8 min maximal running on a treadmill. For F(E)CO(2) and VCO(2) the metabolic cart gave readings that were 2.6% and 1.8% higher and for F(E)O(2) 0.2% lower than the Douglas bag method ( P<0.05). For the metabolic cart and the Douglas bag method the coefficient of variation (CV) for repeated determinations of VO(2) was 1.9% and 1.8%, respectively. For VO(2) and V(E), there were no significant differences between the two methods and the 95% confidence interval of the difference in VO(2) was within -30 and +20 ml min(-1). The CV of the differences in VO(2) between the two systems was 2.4% and it is concluded that a metabolic cart method based on inspiratory flow rate is suitable for measurement of VO(2) and V(E) during both submaximal and maximal exercise.