In this paper, we report the significant role played by the testing chamber for measuring the true gas sensor responses. Two different designs were studied and compared. The first one has a cross-shape in which the direction of gas flow is perpendicular to the sensor surface. The second one has a boat-shape in which the gas flow is tangential to the sensor surface. The results demonstrated that contrarily to the cross-shape chamber, the gas concentration is highly homogenous and equal to the set point in the boat-shape chamber. Additionally, the sensor responses are much stable, faster and higher. Besides, all the mathematical modeling and simulation results are in a good agreement with the experimental studies, which allow the validation of the proposed boat-shape chamber model.