Aiming at the problem that the intelligent recognition algorithm is too complicated in the fault diagnosis technology of circuit breaker, a fault diagnosis method of frametype circuit breaker is proposed based on grey correlation degree. This method uses the grey comprehensive correlation degree to describe the geometric similarity among the fault characteristic sequence curves to judge the type of the fault and makes full use of the changing trend of the fault feature itself. Firstly, the method extracts the local mean decomposition energy moment of the contact vibration signal or the ensemble empirical mode decomposition energy moment of switching coil current signal, which are normalized to form the energy moment sequence. Secondly, the grey comprehensive correlation analysis between the resulting energy moment sequence and the referring sequences under different fault states. Finally, according to the principle of maximum correlation degree, the fault type is determined, and the weight coefficients of various indices of the grey comprehensive correlation degree are obtained with entropy weight method. The diagnosis and testing on the switching mechanical structure fault of the circuit breaker contact and the switching coil circuit fault of operating accessory were conducted, and the results show that the proposed method can achieve the diagnosis of different faults for the frametype circuit breaker accurately and effectively.