Abstract: Remote sensing (RS) is a good tool to estimate wheat yield. In order to get wheat yield estimation model under different nitrogen and wheat stripe rust levels, this paper analyzed the correlation relationship between vegetation index, the first derivative and the wheat leaf nitrogen content, wheat stripe rust disease index(DI), yield components and wheat yield respectively, using regression statistics method to choose high correlation factors to estimate wheat yield. The results showed that the yield estimation model in filling stage used respectively the vegetation index of green red ratio (GR) and the sum of green band first derivative, the sum of blue band first derivative had a good estimation result. Yield models estimation accuracy in 2010 could reach 99.87% and 99.98%, in 2011 could reach 97.9% and 95%. This research found that the hyperspectral remote sensing technology had a good yield estimation result under the stress of nitrogen and wheat stripe rust, the research result discussed in this paper was of great significance to study wheat yield estimation models under more stress and cultural practices.