In view of inter-turn short circuit fault detection of dry air core reactor，a kind of inter-turn short circuit faults detection method using the magnetic field amplitude and the support vector machine is proposed and its effectiveness is verified by simulation. Firstly，the field circuit coupling model for a dry type air-core reactor is set up，the variation in the amplitude and angle distribution of the bottom magnetic field of the reactor under different fault conditions is studied and analyzed and the the variation information of magnetic field amplitude and angle are taken as the characteristic data，the fault classification and diagnosis model based on SSA-SVM is proposed. To consider the limitations in practical applications，the magnetic field distribution is represented by a vector which is composed of magnetic field data at uniform measurement points，and the Euclidean distance is introduced as a mathematical indicator to describe its variation. The simulation results show that the magnetic field information can be effectively applied to the SVM module for classification and diagnosis of inter-turn short circuit faults，and the obtained SSA-SVM model achieves better diagnostic accuracy after optimizing its penalty factor and kernel parameters through SSA.