To explore the effectiveness of Temperature-Vegetation Angel Index in forecasting crop yield with MODIS, winter wheat yield forecast was taken as an example in Shijiazhuang and Xingtai city, Hebei province of China. Firstly, according to winter wheat biological characteristic, the four parameters of normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), temperature-vegetation angel index (TVA) and enhanced temperature-vegetation angel index (ETVA) were calculated in heading stage. Secondly, the regressive models between ground-based measurements of winter wheat yield and NDVI, EVI, TVA and ETVA data from MODIS were established. It was found that there was significantly linear regressive relationship between yield and NDVI, EVI, TVA and ETVA. The correlation coefficient R2 between yield and NDVI, EVI, TVA and ETVA were 0.61、0.65、0.68 and 0.74, respectively. The models based on TVA and ETVA were better than that based on NDVI and EVI, especially the ETVA. This research shows that TVA/ETVA, which integrated vegetation index and land surface temperature, can be applied in yield forecast with improved forecast accuracy.