The invention relates to a fault determination method based on redundancy inertial measurement information, wherein the redundancy is expressed by N which is not less than 3. The method comprises thefollowing steps of: (1) converting a pulse outputted by the current period of N paths of inertial measuring apparatuses to N paths of physical quantity information Txi, wherein i equals to 1, 2, untilN; (2) subtracting the physical quantity information Txi from Txj, and comparing the obtained difference value with a preset threshold sigma x; if the difference value exceeds the preset threshold sigma x, setting a mark delta theta ij to be 1, or the delta theta ij to be 0, wherein i equals to 1, 2, until N, j equals to 1, 2, until N, and j is not equal to i; (3) adding all the marks delta thetaij associated with the ith path to obtain FTxi, if the FTxi is more than or equal to a preset threshold M and not equal to the minimum value in the FTx1, FTx2, until FTxN, showing that the ith path of the physical quantity information TXi has an error, namely, the ith path of the inertial measuring apparatus has a fault; if not, regarding that the ith path of the inertial measuring apparatus hasno fault, wherein the preset threshold M ranges from 2 to N-1; and (4) waiting for an output pulse of the next period of the N paths of inertial measuring apparatuses, executing from the step (1) circularly.