As the installed capacity of wind turbines grows rapidly and cumulative operation time continues extending, the maintenance issue of the wind turbines becomes increasingly prominent, it is necessary to develop effective wind turbine fault diagnosis and prognosis systems urgently. In this paper, the main fault characteristics of wind turbines are summarized from two aspects of fault diagnosis and fault prognosis. Aiming at the difficult problems in fault diagnosis, the research status of the fault diagnosis approaches based on vibration, electric signal analysis and pattern recognition algorithms for wind turbine fault diagnosis are analyzed and summarized. The technical characteristics, limitations and future development trends of different approaches are pointed out. Aiming at various characteristics of mechanical structure and electronic system degradation in wind turbines, current research development of the fault prognostic approaches for wind turbines are summarized. The fault prognostic approach fusing the physicsoffailure model and datadriven model is proposed. Finally, the new development and the problems requiring further study for the fault diagnosis and prognosis of wind turbines using supervisory control and data acquisition (SCADA) data are summarized.