Water is one of the essential natural resources for the development of life on the earth. In this study we apply Disjunctive Kriging (DK) and Radial Basis Functions (RBF) for zoning of groundwater levels. In study area the groundwater levels data have high skewness. Due to samples unsuitable distribution, data was normalized using logarithmic and QQPlot methods. Also geostatistical different methods were evaluated using cross-validation technique. Results showed that Disjunctive Kriging (DK) compared to Radial Basis Function (RBF) has the higher accuracy and the best model of Semivariogramis Exponential model. Also the groundwater levels decreases from north to south of the Shahrekord plain, Iran. Finally, Disjunctive Kriging was selected as the most appropriate method of investigation for the groundwater levels zoning Sahrekord plain.