Concerning the problem of underwater bearings-only system target tracking with incomplete measurements when the probability of sensor detection is less than 1,an improved extended Kalman filtering algorithm for target state estimation was presented. First, the mathematical model of underwater bearings-only system for target tracking with incomplete measurements was established. Second, based on the sensor's incomplete measurement data, the previous update data was used to compensate for the incomplete date and then to perform the filtering. Finally, two evaluation criteria including Cramer-Rao Low Bound (CRLB) and Root Mean Square Errors (RMSE) were used to evaluate the proposed algorithm. The simulation results show that the proposed extended Kalman filtering algorithm for target tracking has higher real-time property with desired tracking precision in the problem of underwater bearings-only system target tracking with incomplete measurements.