For obtaining economically some state value of vehicle which are required by control system, an indirect measuring method was proposed by using the extended Kalman filter technique. The specific flow for the application of extended Kalman filtering theory in estimation of vehicle state parameters was mainly presented, and a 2 degree-offreedom observer model on vehicle dynamics was established based on the theoretic GIM tire model. Through simulating and comparing with a high-precision vehicle dynamics model, the results indicate that this type of soft sensor, which mainly depend on the Kalman filtering technique and some measured easily state information of vehicle such as longitudinal, lateral acceleration, is accurate and effective. It can real-time estimate state variables required by vehicle control system. Therefore, it has very important meaning in reducing the cost of control system of vehicle and promoting the use of active control system in middle and low grade vehicle