A coordination trajectory planning method of multiple unmanned air vehicles (UAV) based on hierarchy strategy is researched, the research is divided in three hierarchies, i.e. route planning, coordination control and trajectory control. Multiple routes for each UAV are given based on K-means clustering algorithm and genetic algorithm. A new method is proposed to solve the problem that the use of the traditional algorithm could not find out the result of coordination variable. According to the six degree of freedom model and the coordination variable, a fixed terminal time optimal trajectory control model is built up for each UAV, thus the optimal control problem is converted to NLP problem with the LPM, and the model is solved by using CFSQP method, and then the plan of control parameters and attitude is realized. The simulation shows that the coordination trajectory of UAV obtained by this method is better in operability, less in computational complexity and higher in computational efficiency, and simultaneously the control command obtained is smooth and easy to operate.