Researching the estimating algorithm for the vehicle yaw rate is providing complement to the sensor measure and reference of the sensor diagnosis. If the estimating precision is good enough, the yaw rate sensor will be canceled and the cost of VDC system will be cut down. Based on the eight-DOF vehicle model and a modified Dugoff tire model which is simple and accurate, the estimation algorithm of vehicle yaw rate is proposed in this paper. The state feedback observer for yaw rate is designed and compared to the observer based on Kalman filter. The simulation results indicate that this algorithm can calculate vehicle yaw rate in real-time and the estimating precision is better than the Kalman filter in nonlinear condition. Therefore, the state feedback observer in this paper is proposing a low-cost and more practical idea for estimating the vehicle yaw rate on-line.