To model the spatial variability of groundwater chemical characteristics, our research was performed using Geostatistic and deterministic methods. In our study, Na+, CI-, HCO3-2, total Cations, TH, SAR, EC and TDS from ground water characteristics were selected. At first, normality of data tested using Kolmogorov-Smirnov test, then we transformed EC, total Cations and TDS using logarithmic transformation. Variography and interaction variogram of data calculated. The result of estimated and expected amount using MAE and MBE were presented. Result showed significance of Geostatistical methods in comparison to deterministic methods and Cokriging increased result's precision. Cokriging is the best method for modeling of Cl-, total Cations and TH. Disjunctive Kriging is the suitable methods for modeling of HCO3-2 and Na+. Universal Kriging resulted best model for TDS. On the other hand ordinary Kriging is the best method for SAR modeling.