The wind data obtained from an AWS(Automated Weather Station) was used to predict the AEP(annual energy production) of Gangwon wind farm having a total capacity of 98 MWin Korea. A wind energy prediction program based on the Reynolds averaged Navier-Stokes equation was used. Predictions were made for three consecutive years starting from 2007 and the results were compared with the actual AEPs presented in the CDM (Clean Development Mechanism) monitoring report of the wind farm. The results from the prediction program were close to the actual AEPs and the errors were within 7.8%.