Abstract: The gravity wave drag triggered by the Tibetan Plateau remains unclear at present. To address the problem, a parameterization scheme for subgrid-scale orographic gravity wave drag was introduced into the GRAPES Meso and a suite of numerical experiments were conducted. Several conclusions from the results are as follows. (1) According to the vertical distribution of orographic gravity wave drag along 30°N, the blocking drag mainly exists in the lower levels (from level 1 to 5) while the gravity wave drag mainly exists between level 5 and level 10. According to the horizontal distribution, the blocking drag, which is dominant on level 3, mainly exists in the flanks of the Tibetan Plateau. Large values of blocking drag are located from the eastern Tibetan Plateau to northern Yunnan-Guizhou Plateau. (2) Analysis of the Froude number and the altitude of circumfluent flow show a large gradability, and thus the highest altitude of circumfluent flow is located at the area of Himalayas and eastern flank of the Tibetan Plateau. The larger the Froude number is over a specific area, the higher the altitude of circumfluent flow is in the area. (3) With the adoption of the subgrid-scale orographic gravity wave drag parameterization scheme, the model results reflect more accurate representation of the breaks of orographic gravity waves in lower and higher levels, as well as its upward transport. (4) Moreover, both the single case study and the batch experiments show positive impacts on the simulation of wind field and precipitation, which leads to the improvement of model prediction accuracy.