This paper studied the fixed-wing unmanned aerial vehicle (UAV) flocking problem from the biological point of view on the basis of the UAV model governed by the complete 12 variables. A weighted and undirected graph is applied to describe the time-variant metric interaction relationship among fixed-wing UAVs. Based on the proposed model and the communication mechanism, a distributed cooperation approach is designed to force groups of fixed-wing UAVs to collaboratively accomplish predefined tasks such as imitating a flock of birds. During the evolution process, four constraint conditions should be considered. The first one is that each UAV flies under bounded state variables, including attitude angle, velocity, and attitude angle speed. Second, one forward speed is necessary for the flight of each fixed-wing UAV. The third one concerns the aviation safety problem. Considering the real size of the fixed-wing UAV, the minimal distance between any two UAVs during the evolution process should be large enough to avoid collisions. The last constraint condition is that the lesser the adjustment time, the more likely it will be within a steady-state error margin. Four constraint conditions are skillfully taken as evaluation criteria to determine the coupling strength of the communication network. Numerical simulations are provided to validate the feasibility of the proposed approach for fixed-wing UAVs to implement the steady flight task of variable-altitude flocking and the steady flight task of invariable-altitude flocking.