A blow-down wind tunnel is a typical nonlinear time-varying system facing the coupling effects between the pressure and temperature during the short-time test procedure. The control of blow-down wind tunnels has been discussed for a long time and a satisfactory general solution to this problem is still of interests. This paper aims to model the internal relationship of the state variables of the wind tunnel by using thermodynamic theories. With the developed model, a new control method combining extended Kalman filter (EKF) together with auxiliary nonlinear predictive filter (NPF) is proposed to improve the control performance of the blow-down wind tunnel controller, in terms of accuracy and robustness. The transient coupling effects between the pressure and temperature are fully considered in the proposed approach. The results from the simulation and experiments are consistent and demonstrate that the controller based on EKF combined together with NPF can work better than previously proposed EKF-based controller.