A study was conducted between 2000 and 2004 on fields in North Dakota, Minnesota and Montana to determine viable methods for delineation of nitrogen management zones. A variety of strategies were used to gather data relevant to delineation, including aerial photographs, satellite imagery, soil electrical conductivity sensors, topography and yield mapping. Three different methods of zone delineation methods were used to form zone boundaries; k-means clustering, neural network clustering and a weighted, classified delineation method. All three methods were effective in forming useful zones to direct variable N rates. Topography, yield mapping and imagery were particularly effective in providing data for the classified method. Economic benefits were achieved in some fields through the use of zone delineation and variable rate N application.