Data from seven Management Systems Evaluation Areas (MSEA) were used to test the sensitivity of a leaching model, PRZM-2, to a variety of hydrologic settings common in the Midwest. Results of long-term simulations using regional and generalized input parameters produced ranks of leaching potential similar to those based on measurements. Atrazine leaching was simulated because the use of atrazine was prevalent in the MSEA studies and it frequently occurs in the region's groundwater. Short-term simulations used sit-specific soil and chemical coefficients. Generalized simulations used data available from regional soil databases and standardized coefficients. Accurate short-term simulations were precluded by lack of antecedent atrazine concentrations in the soil profile and water, raising questions about simulations using data for less than five years. The ranks of simulated atrazine detections among the study sites were similar to the ranks of observed atrazine detection frequencies. Generalized simulations underestimated the occurrence of atrazine for the three sites where very small concentrations were observed. Simulations overestimated concentrations for the four sites where very small concentrations were observed. Simulations overestimated concentrations for the four sites with the greatest frequencies of atrazine detection. Simulations with generalized soils data produced concentrations that compared more favorably with observed detection frequencies than did simulations with site-specific soils data. This is encouraging for regional modeling efforts because soil parameters are among the most critical for operating PRZM-2 and many other leaching models.