Aeroengine gas path fault can be regarded as a very significant part in the engine fault category. Therefore, in order to ensure flight safety, it is necessary to avoid gas path faults and eliminate and predict gas path faults in time. With the development of modern computing power and algorithm, the data-driven method has become more and more crucial in gas path fault diagnosis. Based on the above background, this article analyzes in detail the current situation of engine gas path fault diagnosis based on data-driven method, such as clustering, Support Vector Machines, artificial neural networks, deep learning, etc. Based on the above research, the advantages and disadvantages of various methods are discussed, and it also points out the difficulties and key parts of using data-driven method. Finally, the future research and the development trend is prospected.