Abstract: The orographic gravity wave drag parameterization physics in the ECMWF model is introduced into the Global/Regional Assimilation and Prediction System (GRAPES) that filled up the blank of physics in the GRAPES for global middlerange forecast. Orographic static data needed in the gravity wave drag physics are recomputed and validated through comparing with the original static data from the ECMWF model. Numerical experiments have been performed to test the sensitivity of the GRAPES to gravity wave drag parameterization. The results show that as the orographic wave drag is implemented in the GRAPES, the wind field will be changed over the largescale topography region, the part of zonal wind encountered the QinghaiTibet plateau will pass over the highland with the speed slowing down and the other part of the zonal wind will deflect southward and bypass the highland with a circuitous route. The wind field change is more observable with the model run time extended. The more diversiform is terrain structure, the more various are forecasting wind changes. The successive numerical experiments using GRAPES indicate that the usable forecast period of time is prolonged and then forecasting precision of the global geopotential height pattern is enhanced as the orographic gravity wave drag physics is implemented in the GRAPES. At the end, the pilot study of gravity wave drag physics effect on precipitation is done through a weather case simulation analysis. Evidences suggest that introducing of orographic gravity wave drag physics in GRAPES can improve precipitation forecast of the system since the forecasting wind pattern is changed so as to be closer to the real atmosphere state.