Droughts are weather patterns involving prolonged reductions in precipitation that are distinct from normal weather cycles. They can be accompanied by extreme heat. There are three types of drought that affect vegetation : weather, soil,and physiology. In recent years,global climate change has significantly increased the frequency of drought and other extreme weather in China. Severe drought interferes with agricultural production and has caused a sharp decline in the net primary productivity of vegetation. It has decreased the total volume of rivers, dried up lakes, and degraded local environments. Southwestern China has suffered from a long-term drought that began in the autumn of 2009. Precipitation is half of what it was years ago. In this study, meteorological station data was used to analyze the process and magnitude of this drought from 20092011. Then a light-use-efficiency-based model for calculating net primary productivity called GloPEM was used to determine the impact of drought on the net primary productivity of vegetation during this time. The study area included Guizhou, Yunnan, and Sichuan Provinces, the Guangxi Zhuang Autonomous Region, and Chongqing City. The results showed that the drought was severe in Yunnan,Guizhou,northwestern Guangxi,and in southern Sichuan. And the precipitation and moisture index of 2009-2011 were obviously lower than the average of 1980-2011 in southwestern China. In 2009,the precipitation and moisture index declined sharply. In 2010,the precipitation and moisture index returned to near-normal levels. In 2011, the precipitation and moisture index fell to the lowest point in the past 32 years. For light-use-efficiency-based model,the total radiation is an important parameter. In this study, variations in simulated total radiation were closely correlated with observation results (R~2= 0.84,P < 0.01). This drought may have reduced the net primary productivity of vegetation, decreasing plant's ability to create a carbon sink. In the study area, the average of net primary productivity from 20092011 was 12.55 gC m~(-2) a~(-1) lower than the average value from 20012011. It was 0.017 PgC/a in total. This reduced China's total carbon sink by 7.91%. In 2010 alone, this loss reduced China's total carbon sink by 22. 33% for that year. Variations in simulated net primary productivity were closely correlated with observation results (R~2 = 0.64,P<0.01), which indicated that simulation of net primary productivity from GloPEM model was reliable and the parameters of GloPEM model were suitable in southwestern China. From 20012011,variations in net primary productivity were closely correlated with evapotranspiration (R~2 = 0.44, P < 0.05) in southwestern China. From 20092011,variations in net primary productivity and evapotranspiration were synchronized, but variations in precipitation and moisture index were not synchronous with those in net primary productivity or evapotranspiration. Statistical analysis of areas covered and affected by drought from 20092011 confirmed this. Variations of soil moisture levels were closely correlated with net primary productivity (R~2 = 0.25, P < 0.01). This phenomenon might have a relationship with the water conservation function of the ecosystem,which causes a delayed correlation between soil moisture levels and precipitation.