The Constant Volume Sampler (CVS) is often used to dilute automotive exhaust with ambient air for measurement of emissions from vehicles. If the CVS is equipped with a Smooth Approach Orifice (SAO) to measure dilution air flow, the exhaust flow of the vehicle can then be calculated as the difference between dilution and total flow. Calibration of the CVS and SAO is routinely done using ambient air, but recently it has been shown that water vapor has a significant influence on this calibration and thus has this requirement been added to current emission legislation. However, if this is done then the effect of CO2 in exhaust has not been considered. Further on, when using the CVS to measure exhaust flow, only the Critical Flow Venturi (CFV) will be affected by the diluted exhaust gas composition. The measurement of dilution air will still be unaffected and thus will the resulting exhaust flow be influenced by the unbalanced conditions of the CFV compared to the SAO. Even though the effect of gas composition on flow may be small for the CFV, the influence will be much larger on measured exhaust flow since this is calculated from the subtraction of two large numbers. This paper presents how measurements on exhaust flow are influenced by diluted exhaust gas with different composition. The results are then compared with theoretical studies, which show good agreement.