Abstract: In this paper we firstly analyze the correlation between the number of the injured and the factors such as magnitude, intensity, original time, population density, and the secondary disasters in the earthquake events in Yunnan since 1900. Considering the earthquake disasters' distinct regional differences in Yunnan, we set population density, GDP per square kilometer, and geological disaster susceptibility as risk assessment indexes, and by using secondary classification method, analytic hierarchy process, and GIS spatial analysis method, we divide Yunnan into 3 parts: high-risk area, medium-risk area, and low-risk area to illustrate the people's injury risk. Then in these three areas, we successively fit the relationship between the number of the injured and the magnitude, the intensity in meizoseismal area using Least Square method and get two evaluation models. Finally we use the two models to assess the number of the injured people in the 26 historical earthquake events since 2008. Comparing the assessment results with the actual number, we find that the two models are reliable for the fast assessment of the number of the injured in 1 ~ 2 hours after the earthquake.