Abstract: Aeolian sand is widely distributed in western China, characterized by loose, non-cohesive soil and low water content. The static cone penetration test (CPT), as a widely used and high-quality in-situ testing technique, can easily and accurately obtain soil characteristic information in aeolian sand areas. This study aims to efficiently and accurately simulate the large deformation problem during the static cone penetration process of the cone penetration tests. A material point method (MPM) that combines the advantages of the grid and particle methods is adopted to establish a numerical model for the cone penetration tests in sandy soil. The state-dependent Mohr-Coulomb criterion is used to capture the state-dependent characteristics between the void ratio and the mean stress using a relatively simple set of soil parameters. The established MPM numerical simulation model is first compared with a well-tested calibration chamber test, and the accuracy of the established model is verified through the cone penetration resistance and radial stress curves at different depths during penetration. Subsequently, the initial void ratio of the computational domain is changed to simulate the loose state of aeolian sand. Through a series of cone penetration tests in aeolian sand, numerical simulations are conducted to explore the relationship between the cone penetration resistance and the initial void ratio of aeolian sand. The results indicate that the penetration resistance decreases nonlinearly with the increase of the initial void ratio, but the stable value of the penetration resistance in loose sand does not change with the initial void ratio. Therefore, a relative material saving method can be used to select the design parameters in engineering, improving the economic benefits of engineering construction in aeolian sand areas.