Carbon dioxide exhaled by people can be used as a tracer gas for air change measurements in homes. Good mixing of tracer gas with room air is a necessary condition to obtain accurate results. However, the use of fans in dwellings to ensure mixing is inconvenient. The natural room distribution of metabolic CO2 was simulated in laboratory experiments and verified in a field study. The results of the field measurements presented in this paper support the findings of the laboratory study, i.e. that CO2 is well mixed into the room air even though fans were not used. However, spatial variation of the concentration was detected in both studies when constant CO2 emission was obtained. Several critical points of insufficient mixing were identified. If such locations are avoided, additional mixing may not be necessary when using CO2 as a tracer gas in decay and constant emission methods.