Abstarct: The development of precision agriculture is baxsed on soil properties and special management practices. Therefore, knowing the spatial dependency of soil properties and critical level of nutrients in fields is important in order to achieving higher production and better management. In this study, 82 samples of surface soil collected was from Urmia plain and determined some parameters such as, soil pH, organic carbon, calcium carbonate, electrical conductivity, available phosphorus, potassium, iron, zinc and manganese. All results were statistically evaluated in order to assess the correlation between variables, Pearson correlation was used to. The statistical software package SPSS through normal distribution of data was tested. Using the interpolation method, Kriging interpolation was performed using GS + software and the accuracy of these variables was calculated by distribution map. For more applications, production drawings and specifications baxsed on the proposed method the amount of land for wheat crop from West Azerbaijan Research Center for Agriculture and Natural Resources is offering is calculated. According to the map the amounts of phosphorus in the area were probably in sufficient level or more than sufficient level. This result was also observed for potassium, although the amount of available iron, zinc and manganese were low in most areas. Therefore, to increase the yield and quality of wheat crop is necessary to apply some fertilizers containing iron, zinc and manganese. Also, in order to predict the spatial classification of elements phosphorus, potassium, iron, zinc and manganese in this study, neural network is LVQ4a2 used. The results showed that the artificial neural network can be used, due to low detection error in the phase of training and testing in map zoning of soil elements and soil properties.