To overcome the stringent constraints of on-board resources and improve the fault detection capabilities of formation flying satellites,a Riemann distance-based fault detection data selection method applicable to the attitude control system of formation flying satellites is proposed.The Riemann distance measures the difference between the system's measurement information under fault-free and fault conditions.This is used to establish an evaluation in-dex,which assesses the fault detection capability of the formation flying satellite attitude control system.On this basis,a fault detection data selection method is proposed,providing a theoretical foundation for the optimization of the system output information structure.The simulation has shown that the proposed method can effectively en-hance the fault detection capability of the formation flying satellite attitude control system under the condition of limited on-board resources.