The ALMANAC model was validated in a drought-stressed year and in a period from 1989 to 1998 at different sites of Texas to evaluate its ability in simulating maize and sorghum yields at different spatial scales and to extend its application range. There were 11 sites for maize and 8 sites for sorghum in plot-size simulations, and 9 counties for maize and sorghum in county level simulations. The model showed similar accuracy in simulating both plot-size and county level mean grain yields. It could also simulate single-year yields under water-limited climatic conditions for several sites and mean county yields of maize and sorghum, and had small CV values of mean yields for a long-term prediction. The mean error was 8.9% for sorghum and 9.4% for maize in field scale simulations, and was only 2.6% for maize and-0.6% for sorghum in county level mean yield simulations. Crop models often require extensive input data sets to realistically simulate crop growth. The development of such input data sets is difficult for some model users. The soil, weather, and crop parameter data sets developed in this study could be used as the guidelines for model applications in similar climatic regions and on similar soils.