Topography and land use types are key factors which affect soil properties variability on catchment scale in the loess hilly area. Spatial variation of bulk density and soil water were analyzed according to different land use types and different landscape positions, based on data from 82 points in Zhujiagou catchment in Hengshan county. Correlation analyses were carried out between bulk density, soil water and terrain attributes. Finally, terrain attributes and land use types were used to predict bulk density and soil water spatial distribution by multiple-linear regression analysis. There is little difference in bulk density among different land use types, but there is a tendency: shrub land>grassland>terrace farmland>check-dam farmland>woodland>slope farmland. For soil water content, that the tendency is: check-dam farmland>grassland>shrub land>slope farmland>woodland>terrace farmland. The bulk density series for bulk density according to landscape position is: crest>flat>lower slope>upper slope>middle slope. For soil water content the positions order is: flat>middle slope>lower slope>upper slope>crest. Correlation analyses were carried out between bulk density, soil water and terrain indices. It was found that there is positive correlation between bulk density and compound topographic index (CTI); There is a strong negative correlation between soil water and elevation, and a positive correlation between soil water and profile curvature (Kv). The regression model is precise for the soil bulk density, but the variation is rather large and there is a more smoothing effect on the predicted values for soil water.