Spatial variability of soil has an important influence on the structure and function of soil. The spatial distribution of soil physical properties provides basic and useful information relevant to soil management and ecological protection. A typical red beds basin was selected for this study, soil samples at 0-20 cm were taken from 150 locations in the northeast part of Nanxiong Basin, in which GIS and geostatistics were used to analyze the spatial variability of the soil physical properties. The results show that the coefficients of variation of soil bulk density, total porosity and capillary porosity are 9.82%, 4.47%, and 3.72%, respectively, which indicate weak variation. Pearson correlation indicated that soil bulk density was significantly positively correlated with soil moisture and capillary water capacity (p<0.01), with correlation coefficients of 0.85 and 0.91, respectively, but was significantly negatively correlated with total porosity, capillary porosity and non-capillary porosity, with correlation coefficients of 0.82, 0.71 and 0.94, respectively (p<0.01). The spatial distributions of soil physical properties using ordinary kriging (OK) and empirical bayesian kriging (EBK) methods were subjected to comparative analysis. In addition, different cross-validation indicators were applied to assess the performance of different interpolation methods. Cross-validation demonstrated that EBK performed better than OK. And EBK produced smaller regions of predicted soil physical properties than OK, highlighting the necessity of choosing the appropriate methods in studying the spatial distribution of soil properties.