Identifying plant-environment interactions along with remote sensing provides grounds for designing management methods as well as predicting rice yield in different conditions; accordingly, it is very helpful to use vegetation indices for identifying the vegetation and greenness of farms. The regression between the local and high-yield varieties of rice in 2012 and the NDVI, SAVI, LAI, DVI, and RVI indices derived from Landsat 7 in northern Iran indicate the superiority of the NDVI index in the flowering stage of rice. Results show that the coefficient of determination of the fitted model for local and high-yielding varieties is 0.71 and 0.70, respectively, which indicates the good consistency of the results with the regional data. We evaluated the models for the local and high-yielding varieties in crop year 2013 with RMSE of 406 and 272 kg ha^-1 and NRMSE of 12% and 6%, respectively. Moreover, the simulation results show that the yield of the models is well fitted with the observed values; besides, there is high correlation (R>0.80) between the real and predicted yield values. As shown by the investigation of the region’s soil texture, the fine-texture paddy fields have better yield.