In order to improve the tracking precision, real time performance, and reduce communication costs, distributed particle filter is applied to binary wireless sensor network tracking. The node with the biggest signal strength is selected as cluster, and the dynamic clustering is composed of the cluster and all nodes within one hop communication range of the cluster. State estimation are achieved in cluster heads, particles and their weight are transmitted between cluster heads. Therefore the distributed particle filter tracking algorithm for binary wireless sensor is given. Effect on tracking precision caused by number of particle and node is investigated. Simulation results indicate that the number of particle has greater impact on tracking precision than that of the node number. Meanwhile the distributed particle filter has better real-time performance and lower energy consumption than centralized particle filter.