A distributed attitude coordinated tracking controller is proposed based on the backstepping method for spacecraft formation under the directed communication topology. The multilayer neural network is utilized to estimate the nonlinear uncertainties, when only a subset of the follower spacecraft can receive the information of the leader spacecraft. By considering the attitude tracking performance of the spacecraft formation, the Barrier Lyapunov function is constructed to keep the state variables in the constraint interval and achieve the constraints of attitude tracking errors. According to the algebraic graph theory and Lyapunov theory, the proposed controller can guarantee the ultimate boundedness of the attitude tracking errors for the follower spacecraft. Simulation results show the effectiveness of the proposed method.