Abstract: With the construction of large-scale wind farms,the condition monitoring and fault diagnosis of wind turbines have become important research topics.In the early stages,the condition monitoring and fault diagnosis of wind turbines mainly rely on manual inspections.However,as the installed capacity of wind turbines continues to increase,the cost and difficulty of manual inspections have dramatically risen.In recent years,the condition monitoring and fault diagnosis of wind turbines based on datadriven methods have gradually become hot topics.This paper provides a review of related research from the view of operation data types.Firstly,for the supervisory control and data acquisition （SCADA） system of wind turbines,the research status of condition monitoring and fault diagnosis methods based on SCADA data is analyzed from the perspective of monitoring objects.Secondly,based on the vibration data of wind turbine components,the advantages and limitations of various vibration fault feature extraction methods are analyzed and compared.Then,according to the new condition monitoring and fault diagnosis methods based on image data or data-image conversion data,the existing research is discussed and summarized from two aspects of single image diagnosis and data-image conversion evaluation.Finally,the future research directions of condition monitoring and fault diagnosis are prospected.