This study examines the importance of the orographically induced gravity-wave drag (GWD) parameterization in Numerical Weather Prediction (NWP) and seasonal forecast. The National Centers for Environmental Prediction (NCEP) Global Spectral Model (GSM) was used in this study. Four experiments are designed. Simulations without GWD, with a typical GWD forcing in upper layer. Impact of an enhanced low-level GWD parameterization due to asymmetry and convexity of sub-grid scale orography (Kim and Arakawa. 1995) is also investigated, together with a further revised Kim and Arakawa’s method. Short-range forecasts have been examined for winter of 2001 to evaluate impact on the short time forecast. And simulations with 5 member ensembles have been performed for summer of 1997 to examine the effect on the seasonal prediction. The preliminary results showed that the enhanced GWD in lower-troposphere generally improves wind and temperature field for both of seasonal and weather prediction. In low-troposphere, the typical GWD scheme shows less impact than the enhanced GWD and revised method. For wind field, enhanced GWD methods show better pattern correlation. The enhanced GWD schemes decrease winds in downwind region of mountains and make strong positive eddies on west side of India, and decrease eddies to negative in south Pacific. In comparison of two enhanced GWD schemes, the revised method shows better results. Detail analysis of effect of enhanced method will be discussed.