Abstract: Soil moisture is one of the key factors controlling the distribution of latent and sensible heat flux at the interface of the land surface and atmosphere and can affect weather and climate processes at various temporal and spatial scales because of its certain memory properties. In the numerical model, uncertainty in soil hydraulic parameters is one of the primary reasons for uncertainty in soil moisture simulations. Using the land surface modules coupled with the Yin He Global Spectral Model (YHGSM), this paper introduces the VG (van Genuchten) soil water retention curve model and discusses the influence of two different sets of soil hydraulic parameters on the offline simulation of soil moisture and global medium-range numerical weather forecast. Soil type information needed for yielding soil hydraulic parameters is obtained from the Global Soil Dataset for Earth System Models (GSDE). The offline results show that, except for the large deviations in the simulation of permafrost and organic soil, the land surface module of YHGSM can satisfactorily simulate soil moisture in most parts of the world, and the simulation accuracy is similar to that of ERA5 soil moisture reanalysis products. Soil hydraulic parameters have a certain influence on soil moisture, and the strength of this influence is closely related to soil types and local climatic conditions; furthermore, coarse and medium texture soils are more sensitive to model parameters. The results of the global medium-range numerical forecast experiments by altering soil moisture simulations indicate that soil hydraulic parameters not only affect the short-term forecast of near-surface temperature and humidity but may also lead to considerable changes in large-scale circulations after 6 days of forecast. Therefore, optimizing soil hydraulic parameters and improving the ability of soil moisture simulation are very important for the global medium-range numerical forecasting system. Moreover, the ability to simulate time variations of soil moisture may be more important to a numerical prediction system than simulations of soil moisture absolute magnitudes.