Abstract: In recent years, there has been a great deal of interest in the problem of multi-station multi-target tracking. The recent published paper of S. Mori proposed a new kind of theory of multi-target tracking for distributed sensor networks, K. C. Chang generalized Bar-Shalom's famous joint probabilistic data association algorithm to multi-station case. In the present paper, we propose a new distributed target tracking algorithm for multiple stations based on the basic principles of distributed estimation theory. Without loss of generality,we consider the case where two substations and a main station(this wain station may be the same as one of the two sub-stations)coordinate to track a moving target,and the two sub-stations use different coordinate systern. The two sub-stations measure and estimate the states of the targets and send the sub-estimates to the main station according to some communication strategies. The main station combines the two-estimates optimally and obtains better estimates. We start from the basic formula of the problem and obtain the optimal equations for combining the sub-estimates of the sub-stations just using simple algebra. After the final algorithm has been proposed, we simulate it through an example for four different communication strategies. The simulation results show that our algorithm is effective, and giycs the correct estimates quickly even if the sub-estimates of the sub-stations are all wrong.