Öz Degradation of the lands by salinity under arid climate and poor drainage conditions can be inevitable. In the Harran plain total salt affected areas covers 10 % of total irrigated areas which are mainly located in the low lying parts of the plain where elevation ranges from 350 to 400 m. Soil salinity shows high spatial variability which requires intensive sampling and laboratory analyses. Geostatistical techniques such as simple or ordinary kriging can be used in explaining this spatial variability and estimating soil salinity parameters at unvisited locations. On the other hand, new approaches such as hybrid interpolation methods which incorporate secondary variables into primary variables can help improve the estimation. Estimating soil salinity is a vital issue in soil fertility and management. This study evaluated multivariate geostatistical methods such as regression kriging (RK) and kriging with external drift (KED) and compared them with ordinary kriging (OK) for the estimation of soil salinity parameters. Topographical parameters (i.e elevation, slope and topo wetness index (ln (A/tan))) as well as soil EC values at different depths were used as auxiliary variables. Overall results showed that estimation and mapping accuracy may be improved using multivariate geostatistical methods depending upon the power of the relationship between soil salinity parameters and environmental variables used as covariable.