Recently, wooden dwellings in Hokkaido become more airtight. As concern for indoor air quality has grown, so too has the need to measure interzonal airflows in buildings to asses the distribution of outdoor air. These airflows can be evaluated by using the multiple tracer gas techniques. The balances of tracer gases are expressed by a set of differential equations. In such a case, we have to obtain the suitable time T_s, to calculate the interzonal airflows using the above equations and mass balance equations of airflow rates when the tracer gas concentrations were measured in each rooms. In this paper, we have a set of mathematical solutions for changes of gas concentration for 2 Rooms Model and a set of mathematical solutions for changes of gas concentration for 3 Rooms Model. These solutions are approximated by exponential functions according to the number of rooms. 1) In the case of 2 Rooms Model, the approximated functions for changes of gas concentration have two exponential functions. They have the exponent a and the exponent b(a<b). The time constant T_a and T_b are defined like T_a=1/a and T_b=1/b. The suitable time T_s to calculate the airflows is obtained between the time constant T_a and the time constant T_b. Because the beginning time to calculate them has to be set after the time to acheive adequate mixing in rooms. When the time to calculate them become larger than the time constant T_a, the influence of the function of exponent b become negligible. We have to finish the evalution of airflow rates before that. Seventeen multiple tracer gas tests were conducted by using the above method. 2) In the case of 3 Rooms Model, the approximated functions for changes of gas concentration have three exponential functions. At the first step, airflow rates are calculated from 9 tracer gas concentrations. After that, the tracer gas concentrations are calculated at the one sampling interval late by using the approximated functions. The measured gas concentrations are compared with the calculated values at the same time. If the calculated values and the measured values do not agree within a present criterion, the above procedures have to be repeated at the next one sampling interval. If the values agree within the present one each other, the evalution of airflow rates have done. Fourteen multiple tracer gas tests were conducted by using the above method.