Abstract: With the grid sampling method, the surface and subsurface soil moisture of hilly-sloppy lands was determinated in a subtropical soil region in winter and spring. The region includes tea land, upland and forest land. Spatial structures of the soil moisture were analyzed by geostatistics method and classical statistics method. The results showed that the soil moisture of forest land was more abundant than that of tea land and upland, the coefficient of variation of soil moisture in the entire research region was much higher than that of the single land use. The surface and subsurface soil moisture had similar trends of variation in the same season. According to autocorrelation analysis, soil moisture in the entire region had markedly different variation characteristics because of different land use types. Based on strange phenomena from seim-variogram functions，the spatial variation of soil moisture in forest-upland boundary was lower than that in the other places; on the other hand, the soil moisture in the area of tea -upland boundary had no spatial correlation at all. Sample variation was found to have an anisotropy character in the whole region. With respect to influence factors, such complex distribution of soil moisture was explained by different land use types and microtopographty in winter; however, land use was the dominant factor to the variation of soil moisture in spring. The soil moisture of the whole area except tea-upland boundary had good variable and spatial continuity. So it was advisable to apply geostatistics method to study the spatial variability of soil moisture in red hilly region.