Soil salinity is one of the important issues in agriculture that hampers soil fertility and crop production in arid and semi arid regions. Hence, the study of soil salinity is of great importance and using appropriate methods for predicting and modeling soil salinity is essential in order to control and prevent its development. In this study, temporal and spatial changes of soil salinity modeling were carried out using combined Geostatistics and SaltMod computer model. To predict the soil salinity in five and ten years' period of time, SaltMod was applied and Salinity values were determined at 98 points in downstream lands of Hajilarchay dam located at northwest Iran. Simulation results showed that in the root zone within a period of five and ten years the salinity values with amount of less than 6.3 ds/m increase and with higher values decrease. Since the SaltMod model does not have the capability of spatial analysis and mapping, therefore, the GS+ software, which includes different geostatistics methods was used to evaluate and study spatial variations of soil salinity in desired times.,. Salinity values in non-sampling sites were estimated and relevant maps were drawn. The geo- statistical methods used in this research included Kriging, Cokriging and Inverse Weighted Distance. In addition, for comparing these methods, Cross-Validation was applied using three parameters of Mean Absolute Error (