Soil water is the key factor for plant growth and vegetation restoration on the steeper slopes of the Loess Plateau. Soil hydraulic properties govern the transport of water and nutrients in soils. Spatial heterogeneity of the soil saturated hydraulic conductivities (Ks) determines water infiltration and redistribution in the soil profile. In order to accurately simulate soil water movement in the soil profiles occurring in a particular area or region, an accurate understanding of the spatial heterogenic distribution patterns of Ks is needed. Classical statistics and geo-statistical methods were used to study spatial heterogeneity and distribution patterns of Ks of disturbed soil samples taken from 180 soil profiles (0-200 cm), at 20 cm depth increments, on a slope covering an area of 40 m × 350 m in the Wind-Water Erosion Crisscross Region on the Loess Plateau, using a grid sampling method with a grid unit of 10 m ×10 m. Results showed that the mean value of Ks (5.36×10^-3 cm/s)for the 0-20 cm soil layer was greater than the mean values for layers between 20 and 200 cm, and that the values for each soil layer below 20 cm ranged from 4.32×10^-3 to 4.76×10^-3 cm/s. The variation coefficients for each soil layer below 20 cm were similar, ranging from 0.48 to 0.57, which implies that they were all moderately variable. The distribution patterns of kriged Ks values were also similar for the various layers. Therefore, the value of Ks determined for the 20-40 cm soil layer alone could be used to predict soil water movement instead of determining and using those of the deeper layers in order to simulate water movement in the whole soil profile. The semi-variation function of the distribution patterns for Ks at 0-20 cm and at 140-160 cm could be fitted by a spherical and a Gaussian model, respectively, while those at other layers could all be fitted by exponential models. The range value of the semi-variation function at 0-20 cm was 45.9 m, while for each soil layer below 20 cm the range value varied from 13.8 to 22.8 m. This implies that, in order to effectively show the spatial variability of Ks using the least number of samples to accurately characterize the spatial variability of Ks in the soil profile, we could increase the separation distance and decrease the sampling density by taking samples from the 0-20 cm soil layer. However, the separation distance should be decreased and the sample density increased for representative soil layers below 20 cm depth. The semi-variation function model parameter value of C+C₀ at 0-20 cm was 0.154, and these ranged from 0.202 to 0.276 below the 20 cm soil depth. The C/(C+C₀) ratio was 55% for the 0-20 cm soil layer, implying moderate spatial autocorrelation. In contrast, the values of C/(C+C₀) ratio were between 82% and 95% for soil layers below 20 cm depth, which showed a high degree of spatial autocorrelation. The distribution patterns of the values of C +C₀ and C/(C+C₀) ratio both indicate spatial variability characteristics of the Ks values in the soil profiles in the Wind-Water Erosion Crisscross Region on the Loess Plateau. The spatial heterogeneity of the 20-200 cm soil layer was greater than that of the 0-20 cm soil layer. Our research suggests a more efficient way to sample and determine the spatial variability of Ks on slopes, which would also provide a means to derive accurate information about soil water distribution within soil profiles and across local landscapes.