Understanding the impacts of climate change and human activities on vegetation is of great significance to the sustainable development of terrestrial ecosystems. However, most studies focused on the overall impact over a period and rarely examined the time-lag effect of vegetation's response to climatic factors when exploring the driving mechanisms of vegetation dynamics. In this study, we identified key areas driven by either positive or negative human activities and climate change. Taking the three karst provinces of southwest China as the case study area, a Leaf Area Index (LAI)-climate model was constructed by quantifying the time-lag effect. Then the associated residual threshold was calculated to identify the vegetation change areas dominated by human activities and climate change. The results showed that, during the implementation period of ecological restoration projects from 1999 to 2015, positive impact areas of human activities were mainly distributed among the implementation areas of ecological restoration projects, accounting for 5.61% of the total area. For another, the negative impact areas were mainly distributed across the mountainous area of Yunnan Province, accounting for 1.30% of the total area. Karst landform had the greatest influence on the areas dominated by positive human activities, whereas both topography and karst landform significantly affected the areas dominated by negative human activities. Urban development level had the greatest impact on the areas dominated by climate change. The outcomes of this study provided scientific supports for the sustainable development of ecological restoration projects in China's karst region. Display omitted •The time-lag effect was considered to establish the LAI-climate model.•Residual threshold was used to separate the impacts of climate and human activities.•LAI dynamics changed significantly during the two stages of 1983-1998 and 1999-2015.•Dominant impact areas of climate change and human activities were identified.•Karst landform significantly affected the distribution of human positive impact area.