From the perspective of active and passive radar target tracking data fusion， this paper summarizes the measurement error distribution characteristics of discrete space distributed active and passive radars for the same air target through a qualitative analysis method of radar measurement error combing mathematical approximation processing and geometric figure analysis. A novel fusion method for active and passive radar target measurement data is proposed when the angle between the target and the deployment point of the active and passive radar is small. By optimizing the weighting coefficient allocation criteria to fully combine the advantages of the each radar measurement accuracy， the target tracking data is fused in the geosphere coordinate system to improve measurement accuracy， the target fusion data is smoothed using the extended Kalman filter algorithm. The simulation results show that this method can combine the advantages of active and passive radar target measurement accuracy， reduce the measurement error ellipse after target data fusion， and overall improve the measurement accuracy of fused target tracking data.