Efficient planning and management of land uses relies on the ability to estimate the spatial distribution of soil characteristics in a region by accurate and reliable interpolation methods. In this study, the performances of two stochastic interpolation techniques such as inverse distance weighting (IDW) and radial basis function (RBF), and two deterministic methods of ordinary kriging (OK), universal kriging (UK) and simple kriging (SK) were compared to estimate the spatial variation of salinity, pH and exchangeable sodium percentage (ESP) of soils located in Çerikli Irrigation Area covering a total of 14924 ha land. The study area is located 8 km east of Yerköy in Yozgat province of Turkey and extends along the Delice Stream to the point where Delice Stream merges with Kızılırmak River. Soil samples of 0-30, 30-60 and 60-90 cm depths were taken from 113 locations in the study area. Electrical conductivity (EC), pH and ESP of the samples were determined in the laboratory following the standard analysis procedure. The EC values of the soils ranged from 0.65 to 67.1, from 1.07 to 98.80 and from 0.99 to 54.50 dS m-1 at 0-30, 30-60 and 60-90 cm, respectively. The average salinity increased as the soil depth increased. The results indicated that some locations in the study area have severe salinity problems in the soil profile. Evaluations of root-mean-square error (RMSE) and mean absolute error (MAE) revealed that RBF-IM as deterministic and SK as stochastic interpolation method performed better estimation results compared to the other interpolation methods. Therefore, significant errors may occur in the interpolated values using the predicted or default method of the software`s used in the mapping of spatial distributions. For this reason, more than one interpolation method should be achieved in mapping of each soil attribute obtained by depth, and therefore estimations and mapping should be achived by using the most accurate interpolation method.