Abstract: In order to monitor the wheat stripe rust with hyperspectral remote sensing and estimate the wheat yield. This study was to find the high yield component of four different resistance wheat cultivar on stripe rust mixture. The canopy reflectance of the different stripe rust severity of wheat cultivar mixture was obtained by hyperspectral remote sensing. Vegetation index was chosen from the spectral data, the correlations among DI, yield and spectral vegetation index were analyzed, and inversion models were built. The results showed that the yield of A(nongda195):C(0045): D(nongda211) = 1:1:1 was the highest. At milky maturity, the correlation between vegetation index and yield was high. The remote sensing models in different growth stages for yield and yield component were built by NDVI. At grain filling, the model of yield was best, after the occurrence of wheat stripe rust, NDVI of multi-temporal combination simulated yield with higher precision.