Aiming at the phenomenon of the temperature drift faults of sensors in bench test of aero-engine, a method based on Empirical Mode Decomposition(EMD）algorithm and Hilbert transform is proposed for the effective validation of aeroengine sensor datum. Firstly, the failure modes of aeroengine sensors, the EMD theory and the Hilbert transform method are introduced innovatively to the analyzing field of aeroengine sensor signals; Secondly, with the faults of sensor signals being decomposed and extracting their fault features by the method, the severity of sensor fault signals is identified and the original signal failure modes is removed through the failure signals reconstruction of aeroengine sensor. Finally, based on the effective signals of aeroengine sensor faults, the effective validation of aeroengine sensor datum is completed. It is proved by practical calculation and analysis that the method is effective for the data validation of engine sensor in this paper.