Abstract: It is of great practical significance for regional ecological management to understand the quantitative impacts of human activities on vegetation under climate change. Based on GIMMS NDVI3g data, meteorological data (temperature, precipitation) and standardized precipitation evapotranspiration index (SPEI), we used correlation analysis and trend analysis to examine the spatio-temporal variation of vegetation and its driving factors in different periods from 1982 to 2014 in the Beijing-Tianjin sandstorm source region. Regression analysis and residual analysis were used to quantify the impacts of human activities on vegetation changes in different sub-regions. The results showed that from 1982 to 2014, the degradation status in 77.1% of degraded vegetation was significantly improved and 64.1% of vegetation had an increasing trend in the study area, with mean annual NDVI decreasing from southeast to northwest. Vegetation coverage increased in 74.5% of the areas after the implementation of the Beijing-Tianjin sandstorm source control project, with mountains in northern Shanxi showing the most obvious increases. Among all the climate factors, rainfall had the strongest correlation with vegetation change. Human activities, such as ecological engineering, played an active role in most areas, especially in mountains of northern Shanxi, where the contribution of human activities reached 94.9%.