Crop yields are highly variable across field and years as a result of the complex interaction between many factors, including topography, soil properties, management practices and weather conditions. There is a limited information on how agronomic management influence yield variability over undulated field through multiple years. This objective of study is to analyze and quantify effects of management practices on relationship between grain yields and field topography. The data used was from the Long Term Ecological Research (LTER) site at southwest Michigan. The experiment was one-factor complete randomized block design (RCBD) with 6 replicates. The four treatments were chisel plowed with conventional chemical inputs, not-till with conventional chemical inputs, chisel plowed with low chemical input and a winter leguminous cover crop, and organic-based chisel plowed with zero chemical input and a winter leguminous cover crop. Grain yields of corn, soybean, and wheat alternatively planted from 1996 to 2002 were recorded using a commercial yield monitor mounted on a combine with a GPS receiver. Elevation was measured with real time kinematic GPS receiver at about 2.5 x 5 m point. Topographical attributes were derived from the elevation data using surface hydrological analysis using ArcInfo GIS software. Correlation and regression coefficients between yield and topographical attributes were included to study the relationship using analysis of covariance and variogram behavior. Primary results showed that the relationship among treatments differed at P < 0.1 level. The treatment of chisel plowed with conventional chemical input had weakest relationship between yield and topography compared with others. The relationship varied across years at P < 0.01 level.