Abstract Owing to the size limitations of launchers, foldable main wings and foldable vertical tails are usually used in the design of barrel-launched unmanned aerial vehicles (UAVs), and the folding mechanisms are complex and bulky. Further, the coupling effect between aerodynamic performance and control capability is not typically considered in the aerodynamic layout design of traditional drum-launched UAVs; thus, the overall performance is often improved by serial design and multi-turns optimization, which result in low design efficiency. Based on the available research on barrel-launched UAVs, this paper presents an optimal design scheme for non-vertical tail aerodynamic layout considering the control capability of multiple control surfaces. A flying-wing configuration, with two groups of ailerons and one group of all moving tips, is employed in the design, and its control capability and aerodynamic performance are quantitatively described in terms of the volume of attainable moment subset and lift–drag ratio. The focus of the current design is shifted from improvement of single aerodynamic performance to optimization of overall performance. To overcome the computational complexity observed in direct optimization, the number of design variables is reduced by identifying the key test factors. The concept of a surrogate model matrix comprising cell surrogate models is proposed to effectively fit the aerodynamic characteristics and control capability of the wing. A multi-objective optimization based on the surrogate model matrix is carried out using the neighborhood cultivation genetic algorithm, which effectively improves the optimization efficiency. The entire process is implemented on an ISIGHT platform for automatic optimization design of the wing. At two design points selected on the pareto frontier, the results show that the lift–drag ratio is increased by 28.36%, reachable torque space is increased by 64.29% in state 1, lift–drag ratio is increased by 23.11%, and reachable torque space is increased by 164.29% in state 2. These results provide a reference for the aerodynamic layout design of barrel-launched UAVs for practical control requirements.