Abstract: The inherent stochastic volatility of wind power affects operation stability and security of the electric power system. Wind power forecasting is one of above solution, and the forecasting accuracy and computational efficiency affect its application in power system. Therefore, grouping method for wind turbines is presented by considering flow correlation and is applied in wind power forecasting in this paper. Because the traditional earth plane coordinates could not reflect the flow information, a new coordinate system, wind farm prevailing wind coordinate system, is defined to reflect the flow characteristic of a particular wind farm. This coordinate facilitates the combination of flow characteristic with the wind turbines grouping method and wind power forecasting technique. Taking a wind farm in northwest of China as example, the proposed method is validated using GA-BP forecasting model, and the results show that the proposed method can make good use of the characteristics of flow correlation in a wind farm to group wind turbines, and balance the forecasting accuracy and which guarantees the economic operation of power system and wind farm.