In this paper, the three-dimension path planning for unmanned aerial vehicle under moving threat was studied. First, the state information of moving threat was predicted by adaptive Kalman filter. According to the relation between heading angle, flight path angle with line-of-sight angles, a mathematical model for controlling track angle was established and a path planning algorithm was presented based on parameter adjustment. Simulations show that the algorithm can assure unmanned aerial vehicle to avoid the moving threat efficiently and quickly.