An accurate and management sensitive simulation model for tile-drained Midwestern soils is needed to optimize the use of agricultural management practices (e.g., winter cover crops) to reduce nitrate leaching without adversely affecting corn yield. Our objectives were to enhance the Agricultural Production Systems Simulator (APSIM) for tile drainage, test the modified model for several management scenarios, and then predict nitrate leaching with and without winter wheat cover crop. Twelve years of data (1990–2001) from northeast Iowa were used for model testing. Management scenarios included continuous corn and corn–soybean rotations with single or split N applications. For 38 of 44 observations, yearly drain flow was simulated within 50 mm of observed for low drainage (< 100 mm) or within 30% of observed for high drain flow. Corn yield was simulated within 1500 kg/ha for 12 of 24 observations. For 30 of 45 observations yearly nitrate-N loss in tile drains was simulated within 10 kg N/ha for low nitrate-N loss (< 20 kg N/ha) or within 30% of observed for high nitrate-N loss. Several of the poor yield and nitrate-N loss predictions appear related to poor N-uptake simulations. The model accurately predicted greater corn yield under split application (140–190 kg N/ha) compared to single 110 kg N/ha application and higher drainage and nitrate-N loss under continuous corn compared to corn/soybean rotations. A winter wheat cover crop was predicted to reduce nitrate-N loss 38% (341 vs. 537 kg N/ha with and without cover) under 41-years of corn-soybean rotations and 150 kg N/ha applied to corn. These results suggest that the modified APSIM model is a promising tool to help estimate the relative effect of alternative management practices under fluctuating high water tables.