Abstract: This research was conducted to study spatial distribution of soil nutrients including nitrogen, phosphorus and potassium and to compare the efficacy of various geo statistical approaches in the estimation of these nutrients and the preparation of spatial variability maps of these elements aimed to righteous management of fertilizers in Urmia plain, West Azerbaijan Province, Iran. To estimate the rates of soil nutrients including nitrogen, phosphorus and potassium of non-sampled areas, the Kriging, Co-kriging and Weighted Moving Average methods were used in a GIS system, instead. To compare these methods, cross-validation method with two statistical parameters Mean Absolute Error and Mean Bias Error were applied. Present studies implied that Kriging method has the least MAE of 0.450 and MBE of 0.025. This method with correlation coefficient of 0.99% and Gaussian semi variogram was of high preciseness in the estimation of nutrient rates in the points with no former information available. The estimation error in this method ranged between 0.18 and 0.75 and its deviation was between -0.002 to 0.12 meq/100 g of soil.