Abstract: Exploring the evolution of driving forces behind vegetation changes over a long time scale is crucial for implementing ecological governance projects.Based on GIMMS NDVI3g data and ERA5-Land temperature, precipitation, and solar radiation data, the PolyTrend algorithm and improved RESTREND method were used to analyze the spatiotemporal characteristics and driving force evolution of vegetation changes in China during the period from 1983-2022.The results indicate that: In the past 40 years, the overall growth of vegetation in China has been significant; The combined effect of climate change and human activities is the main driving force behind the increase of vegetation in China, while individual human activities are the main driving force behind the decrease of vegetation in China; The pixels with non-linear vegetation changes account for 50.6% of the total vegetation pixels, mainly in the Loess Plateau, North China Plain, Qinghai Tibet Plateau, Sichuan Basin and Northeast China; The types of driving force changes with the highest proportion of vegetation ID,DI,IDI,and DID change areas are DH,HD,DHD,and HDH,respectively.It can be seen that the combined effects of climate change and human activities mainly affect the increasing stage of vegetation nonlinear changes, while individual human activities mainly affect the decreasing stage of vegetation nonlinear changes.The research results suggest that more attention should be paid to the role of human activities in vegetation improvement.