Abstract:0.802。The canopy hyperspectral reflectance data were recorded at the main harm stages of rice infected with Sogatella furcifera. Statistic multianalysis method was conducted on the correlation between canopy reflectance data and the Sogatella furcifera number of per hundred individual rice. Three bands, near infrared band 779nm, 1685nm and visible light band 558nm, were confirmed being sensitive to insect pest of Sogatella furcifera. Combining two bands reflectance (779nm and 558nm) into vegetation indices of RVI (ratio vegetation index) and NDVI (normalized difference vegetation index), five single variables of linear function models against the Sogatella furcefera number were established based on sensitive bands, RVI and NDVI. The results show that the five types of function models have significant correlation if α=0.01. The model established by using of the band 779nm is optimal as using sensitive band and the correlation coefficient is 0.921. The model established by using of RVI is optimal as using vegetation index and the correlation coefficient is 0.802.