EPIC (Erosion Productivity Impact Calculator) is a crop growth model developed by USDA to simulate crop growth for estimating crop yield under given natural conditions and agricultural land practices including crop calendar. By combining EPIC with GIS, spatial distribution of crop yield can be estimated if we have data faithfully representing the spatial distribution of natural conditions (e.g. soil, climate) and agricultural practices. However, it is not always easy to acquire such geographic data, especially data on crop combination and calendar. Although several existing studies suggested that the estimation accuracy can be improved by using such kinds of geographical data, there have been almost no studies which quantitatively report how much improvement is actually achieved by using such geographic data. Through a case study of integrating EPIC with GIS in Yellow River Basin (Mid-stream region) of China, this study revealed that the average estimation errors are reduced from 21.7% to 16.3% for wheat and from 47.2% to 22.3% for maize by using spatially-detailed geographic data on crop combinations and crop calendars. Since crop growth models like EPIC can simulate not only crop growth processes but also some of environmental impact processes such as leaching of nitrogen due to the excess fertilizer applications and soil erosion processes, the results imply that the integration of crop productivity models with GIS can effectively support decisions in striking balance between environmental conservation and agricultural activities in river basin management.