Numerical forecast models are widely used as an important tool of proj ecting global climate change. However, their resolution is too coarse to provide the regional scale information. Therefore, downscaling methods for extracting r egional scale information from output of numerical forecast models are dev eloped. Statistical downscaling techniques can be used to predict regional scale climate by using statistical relationship between the largescale climate and t he regional scale climate. On the basis of introducing the theory and approach o f statistical downscaling, application of statistical downscaling to the medium term weather forecast was simply discussed. 