Abstract: Many studies show the importance of vegetation to climate. Southwest China neighbors to Tibetan Plateau, by which climate of this region is influenced by its own system. Therefore, taking account of both factors it is considered that climate of Southwest China is affected by vegetation over Tibetan Plateau to some extent. The lag relationships between NDVI over Tibetan Plateau and temperature of Southwest China are discussed by using statistical methods. Empirical orthogonal function (EOF) analysis is made first to describe temporal and spacial characteristics of mean temperature, highest temperature and lowest temperature in summer over Southwest China during the period of 1982—2001. The results show that there is a coincidence between mean temperature and highest temperature, especially Sichuan and Chongqing are in a cool period during 1980s and getting warm in 1990s. Based on the monthly Normalized Difference Vegetation Index (NDVI) data that is made by GIMMS (Global Inventory Modeling and Mapping Studies) workgroup and the monthly mean air temperature, maximum air temperature and minimum air temperature data of 96 stations in Southwest China, the relationships between NDVI over Tibetan Plateau and temperature of Southwest China and key regions are analyzed by applying correlation, composite and singular value decomposition (SVD). The results show that mean temperature and highest temperature are more sensitive to vegetation change, and the correlation between temperature and preceding NDVI has a significant difference between tow parts of the Plateau, vegetation over the West Plateau shows a stronger influence on temperatures than the east part. Compared with the years of abnormally low NDVI, abnormally high NDVI in winter over the Plateau corresponds to a warming in Southwest China in summer, highest temperature increases most obviously, the maximum increase occurs in July in north of Southwest China. The best coupled patterns of NDVI and tem peratures are showed respectively, the extent of vegetation's impact on temperatures and key regions are distinguished. The strongest correlation is found between the Plateau NDVI of winter and mean temperature of summer; mean temperature is driven by vegetation in northeast part of the Plateau, Lhasa and Linzhi predominately. For highest temperature the area of remarkable correlation is located in west of the Plateau and the large area along the north edge of the Plateau, and for the lowest temperature is the belt region along the north edge of the Plateau. The regions sensitive to the change of winter NDVI are mainly located in Sichuan Basin and the vicinity of it.