Abstract: The study of temporal vegetation dynamics has been one of focused issues on global change, which has great significance to regional ecological management. Based on time series of MODIS vegetation index data from 2001 to 2014 and the DEM data and land use data in southwest China, the trends of the vegetation changes were calculated and the spatial differentiation characteristics of the changes with the other relative factors were analyzed. The results indicate that 74.52% of change trends of Seasonally Integrated Normalized Difference Vegetation Index (SINDVI) are not significant. The areas with significantly increasing trends account for 22.07% and the areas with significantly decreasing trends account for 3.41% of the total, respectively.So the increasing areas are much more than the decreasing areas. Through analysis results of the terrain factors, the trend of vegetation change in the low or middle altitude areas and gentle slope areas are the most obvious. The trend of vegetation change below the altitude of 3, 500 m is more significant than that above 3, 500 m. With the increase of slope, increasing and decreasing trends are both smaller. From the analysis results of land use, the SINDVI change trend is the most remarkable on the artificial surface, with relatively faster increasing and decreasing trends. With the influence of human activities, the increases of the artificial surface and the bare land and the decreases of the forest land are the main reasons for the vegetation degradation.