As to the issue of identification and elimination of the outliers in data measured by airborne system, an improved polynomial extrapolation fitting algorithm is proposed, and the outliers of GPS and AHRS data are identified and corrected in FPGA.Standard deviations of parameters are calculated out according to statistic characteristics of the data, and the polynomial extrapolation fitting algorithm is improved to make it adaptable for identification and elimination of outliers.The results of experiment show that this method can identify and correct the outliers effectively, and can satisfy the requirement for flight test.