Climate change can affect agricultural production and soil and water conservation. The objective was to simulate the first order sensitivity of water resources and soil erosion to assumed climate changes in the region near El Reno, Oklahoma using the Water Erosion Prediction Project (WEPP) model and a climate generator (CLIGEN). Monthly forecasts for the periods of 1950-1999 and 2070-2099 for the region, projected by the UK Hadley Centre's third generation general circulation model (HadCM3), were used. Three emissions scenarios (A2a, B2a, and GGa1) were selected. The WEPP model was run for each scenario in three tillage systems. Projected changes in monthly mean and variance between the two periods at the HadCM3 grid scale were downscaled to daily climate data using CLIGEN. The HadCM3 predicted a general decrease in precipitation in the study area over the century. As a result of precipitation reduction, predicted plant evaporation, soil evaporation, and the long-term soil water reserve were reduced in all three climate change scenarios. However, the decreases in rainfall did not necessarily result in reductions in wheat yields, largely because most decreases were predicted in the non-growing season. About a 1% decrease in wheat yields were predicted for A2a and B2a, while a 6 % increase was predicted for GGa1. Regarding soil erosion, the results indicate a possibility for increasing erosion despite the predicted decreases in annual precipitation due to the predicted increases in precipitation variability. This is especially true for GGa1, in which soil loss was predicted to increase about 20 to 50%. However, simulated results suggest the no-till system is sufficient to maintain low erosion levels in the regions and thus protect the soil and ecosystem under the assumed climate changes.