Accurate and timely estimation of crop yields is crucial for agricultural planning, food security, and economic forecasting. Remote sensing-based models have proven effective tools for large-scale crop yield prediction. This study investigates the transferability and robustness of a remote sensing-driven crop yield estimation model across diverse geographical regions within the U.S. Midwest. Originally developed for Iowa, the crop yield estimation model was adapted for use in Illinois while retaining its original structure and coefficients. The model integrates time-series Normalized Difference Vegetation Index (NDVI) data from Landsat 8–9, with surface parameters including land surface temperature, soil moisture, and precipitation. The study focused on corn and soybeans, the predominant crops in the region. The adapted model projected average yields of 280 bushels per acre for corn and 50 bushels per acre for soybeans across Illinois. These projections align closely with historical yield data and fall within the expected ranges for the state. Spatial analysis identified yield patterns that align with known variations in agricultural productivity variations across Illinois, particularly showing higher yields in the fertile central areas. Our findings have significant implications for large-scale crop yield forecasting and agricultural decision-making. The model’s demonstrated transferability provides a cost-effective method for estimating yields across vast geographical areas, eliminating the necessity for extensive local calibration.