Aimed at the problems that the noise components are complex and the fault features in fiber optic gyro output signals are hard to extract, a fault diagnosis method based on improved sparrow search algorithm and support vector machine (SVM) is proposed. Firstly, the normal and the fault signals of fiber optic gyro are decomposed by threelayer wavelet packet to extract feature vectors. Secondly, by introducing the improved Logistic chaotic map and adaptive t distribution strategy, and adding the boundary exploration and alarm cancellation mechanism, the ISSA is proposed and applied to the parameter optimization of SVM. Finally, a SVM model is established for fault identification and diagnosis of fiber optic gyro. The results show that the method proposed in this paper can be effectively used in fiber optic gyro fault diagnosis, and simultaneously effectively improve the accuracy of fiber optic gyro fault diagnosis compared with the sparrow search algorithm, the gray wolf optimization, the particle swarm optimization, the genetic algorithm and the Beetle Antennae Search.