Groundwater aquifers as one of the main sources of water is faced with different risks such as loss, reduced food due to decreased rainfall and natural and non-natural contamination. Thus, monitoring of the quality of groundwater resources and water quality analysis is one of the important parts of groundwater studies. The aim of this paper was to model the spatial variability of chemical properties in groundwater using geostastical methods in Bajestan plain in located Razavi Khorasan Province. Hence, 61 water samples were collected from different wells of the plain and some parameters of groundwater such as Na, SAR, SO4, EC, Cl, TDS, HCO3, pH, Ca and Mg were selected and were evaluated. First, normality of the data was tested and Log -transformed was applied to some parameters in order to make its frequency distribution more normal. The analysis of semivariograms and cross-semivariograms of different parameters were calculated. After selecting the optimal semiovarigrams based on statistical amounts of RSS and R2, some candidate gridding methods including Kriging, Cokriging and Inverse distance weighted (IDW) were examined to determine the best method for creating the interpolation maps and the quality of that were tested by Cross-validation. The results showed that geostatistical methods of Kriging and Cokriging had high ability for interpolation and estimation of spatial variability of selected parameters compared to the IDW method. Also, results indicated that best modeling method for the parameters of Mg, Ca, EC, SAR, so4 and Cl and other parameters such as Na, TDS, HCO3 and pH was Cokriging and Kriging method, respectively. On the other hand, the quality of water wells was examined for drinking and agricultural utilization. The results showed that the salinity and sodium adsorption ratio was high in most of the wells. Thus uncontrolled utilization of them for long time caused salinity and reduce soil permeability and intended for drinking are caused the deposition and corrosion in pipes.