Abstract: A three-dimensional randomly packed granular bed with a height of 20 mm and a granular diameter of 1.0 mm was simulated on the basis by coupling of computational fluid dynamics (CFD) and discrete element method (DEM) for liquid-solid phase filtration. Fluid flow and particle deposition characteristics were analyzed, and effects of deposition growth on the filtration efficiency and pressure drop were investigated as well. The results show that with deepening of the filter bed, the initial pressure drop is gradually increased. Compared with the values from Ergun equation, the calculation deviations of the initial pressure drop at different fluid flow rates are within 25%. The initial filtration efficiency for larger particles is higher, and it decreases with the velocity of fluid flow. On the basis of regression analysis, it can be predicted that the deviations between the predicted and the experimental values are within 10%. In addition, the particle deposition fraction in the filter bed decreases with the depth of the filter bed, and particles near the inlet incline to form clusters, and their proportion at the early stage of filtration becomes prominent. When the filtration goes gradually deepening, the numbers of particles that accumulated along the bed become more and more even. The deposition of particles influences the flow of fluid within the filter column, which subsequently affects the filtration efficiency and pressure drop. For the particles in different sizes, the deposition growth on the granular surface leads to an increasing trend of filtration efficiency and pressure drop.