With the development of wind power and scale of wind farm, the spatial distribution difference between wind turbines also increase. Besides, wind turbine trip-off and damage accidents has occurred frequently because of the severe wind conditions, having adverse impacts on the stability and safety operation of power grid. Therefore, it is necessary to study the online risk assessment method for power system with wind energy. Considering the wind turbine spatial distribution difference, this paper proposed an online disturbance risk measure of wind farm based on decision trees, which can perform data mining on online information, and make fast judgement on voltage violation and wind turbine trip-off. Furthermore, according to the judgement of decision trees, disturbance risk measure indices are proposed, which are visualized and provide supportive information for wind farm and power system operators. © 2015, The editorial office of Transaction of China Electrotechnical Society. All right reserved.