Abstract: To solve nonlinear, non-Gaussian system problems in target tracking, the paper focuses on Bayesian filters and its realization, the bootstrap filter (BSF), which is compared with the extended Kalman filter (EKF) as to the accuracy, computational load, and other aspects. A simulation of a typical model is presented to compare the performances of BSF and EKF. The simulation shows that the performance of BSF is superior to that of EKF.