Wetland vegetation biomass is a significant indicator of the health condition of wetland ecosystem, which directly mirrors the growth and productivity of vegetation communities.The estimation of vegetation biomass in alpine wetland contribute to understanding the causal feedback relations between alpine swamp ecosystem and global climate change.Longbaotan nature reserve in Sanjiangyuan area was taken as the research region.Taking advantage of hyper spectral dimension and multi-angle stereoscopic structure information of CHRIS/PROBA,we analyzed the correlation between wetland vegetation biomass and some remote sensing factors,including spectral reflectivity,narrow band vegetation indices,red edge indices and principal component.The angular sensitivity of biomass was explored at the same time.The results showed that the biomass of vegetation on the alpine wetland was sensitive to satellite observation angle.The forward observation of the+36°degree image was significantly better than 0 and-36°degree images.Besides, the exponential model was the best regression model in the nonlinear models.Among them,REIP index of-36°degree was the fittest variable to biomass,which R~2 was 0.599 and Fvalue was 37.404 in dry biomass,and R~2 was 0.685 and Fvalue was 54.410 in fresh biomass.The maximum of dry biomass in Longbaotan area is 446.7 g/m~2,and fresh biomass is 2 368.1 g/m~2.