Abstract: An Extended Kalman Particle Filter (EKPF) integrated with negative information (scans or dwells with no measurements) is implemented for target tracking in the Stand-Off Jammer (SOJ). In the EKPF, the Gaussian sum likelihood function which is derived from a sensor model accounting for both the positive information and negative information is directly used in the weight update of the particle filter. And the importance density function is generated by using the Extended Kalman Filter (EKF) to take full account of the current measurement, thus leading to the distribution of the particles approaching the posterior probability density function. Moreover, use of a small number of particles can achieve good tracking accuracy. Simulation results show that EKPF outperforms the EKF implementation in terms of track continuity and track accuracy but at the cost of large computation complexity.