As it is the case for soil chemical and physical properties, greenhouse gas fluxes also exhibit tremendous variability across the field. However, because of the cost involved with collecting numerous samples, measurements of fluxes across agricultural fields are often limited to few points. The average value of point measurements is later used to predict the total flux for the sampled area or the entire field. This approach may result in an over or underestimation of the total flux. The objective of this study was to assess how using geostatistics, geographic information systems and map analysis could improve the estimation of N2O, CH4 and CO2 fluxes from an agricultural field. The study was conducted at Lincoln University's freeman farm in corn and soybean fields. The approach consisted in sampling N2O, CH4 and CO2 fluxes at 32 locations, fitting a variogram model to fluxes data with GS+, predicting fluxes at un-sampled locations by kriging when the data fitted into a model, mapping the entire field with ARCGIS 9.2 and classifying map zones with Multispec 3.2 and finally displaying classified maps. After classification, an improved total field flux for each of N2O, CH4 and CO2 was calculated by a summation of all zones multiplied by their corresponding flux values. This improved total field flux was compared to the total flux computed using the field average flux value, for each greenhouse gas. Results obtained showed that the total N2O, CH4 and CO2 flux calculated using a combination geostatistics, geographic information systems and image analysis was 9.0 to 200% higher as compared to that calculated using the field average flux. Our approach seems to provide in an improved quantification of greenhouse flux. The approach can be extended to other soil and environmental parameters.