A concurrent subspace optimization method is developed for the layout design of wings. In terms of the type of variables, the design problem of structures is decomposed into three concurrent subspaces with spar location, stringer and thickness optimization. The optimization in each subspace is carried out separately to avoid the coupling of design variables. By remaining the design variables in the lightest subspace unchanged and altering two other types of variables with approximate one-dimensional search algorithm, the coordination is achieved in system level to ensure the convergence of the analysis problem. Then next iteration is executed until termination. An UAV example is solved successfully to demonstrate that the proposed method is feasible and has practical engineering value.