The invention discloses an engine fault diagnosis method based on a BP neural network. The method comprises the following steps: (1) collecting engine fault data, and listing out engine fault reasons;(2) determining optimal hidden layer node number of a BP neural network model, and establishing a BP neural network model; (3) training the BP neural network model according to the existing fault data; and (4) analyzing the analyzed engine data by utilizing the BP neural network model obtained through the training, and determining the fault reason corresponding to the data. Formerly, the engine fault diagnosis has the defects of being complex in mechanism, low in detection precision, high in cost and incapable of displaying the fault reason, the diagnosis method is mainly applied to the faultdiagnosis aspect of the engine, and has advantage in comparison with the previous method, the cost is saved, the modelling efficiency is improved, and the optimal hidden layer node number can be quickly locked.