The recent trend of expanding the cropping area of corn silage (Zea mays L.) in Konsen district, Hokkaido, the largest dairy production area in Japan, has raised concerns about increased nitrogen (N) leaching and groundwater pollution by nitrate. This study aimed to evaluate the applicability of modified LEACHM (Leaching Estimation and Chemistry Model), an improved process-based numerical model for simulating N dynamics, particularly in Andisols, for predicting N leaching in a corn field in the Konsen district. The study area has a lower temperature and higher humidity than the average in Japan. Corn silage was grown yearly from 2007 to 2012 with three different fertilization treatments: one experimental plot was treated with chemical fertilizer (C) and the other two plots with organic fertilizers of composted cow manure (M) and slurry (S), respectively. A monolith-type capillary lysimeter packed with an undisturbed subsurface soil core of 20–70-cm depth and a diameter of 25 cm was buried below the plow layer in the field to measure leachate volume and inorganic N concentration. The simulated results using modified LEACHM showed a reasonably good agreement with the measured inorganic N concentration patterns, such as the magnitude and timing of peak N concentrations, in the three treatments despite the differences in the applied N sources. However, the performance of the model with respect to cumulative amounts of inorganic N leached yearly and/or during the whole experimental period varied between years and plots. Overall, the model was a valuable tool for predicting N leaching and examining various scenarios in corn silage fields treated with different N sources in this region. Future studies on separately determining the N mineralization rates of humus and crop residue, as well as the N transformation of composted manure and slurry, would be valuable in further improving the model’s prediction.