Due to the significant climatic effects of snow cover, it is very important to improve numerical simulation of snow cover in climate models. With the field data from Col de Porte, France and BOREAS, the evolution of seasonal snow cover is simulated with Comprehensive Land Surface Model (CLSM). The objective of this study is to validate the model and investigate the snow cover processes in both the alpine regions and the boreal forest. The sensitivity of the simulated results to some model parameters and the different physics responsible for the snow cover variation in vegetated and non-vegetated cases are investigated. The modeling results are in good agreements with the observational, and the model represents the snow-pack development and both the timing and the rate of seasonal snowmelt accurately in both cases, indicating that the model has the capability to capture the main features of seasonal snow cover under different conditions. Results also suggest that different parameterization schemes of the snowfall density and snow water holding capacity have significant effects on the simulation of snow cover. The physical processes related to the snow cover variation are different whether vegetation exits or riot. Vegetation plays an important role in the dynamics of seasonal snow cover by controlling the radiative fluxes at the snow surface and thus the surface energy balance.