A typical slope in the Caijiachuan watershed, Loess Plateau of China was an object of research.A total of 313 soil moisture monitoring points were laid out in a 20 m×20 m grid pattern in the experiment site for measuring soil moisture in two layers, 0~30 cm and 30~60 cm, with TDR (time domain reflectrometry) in 2005.Based on the data, a soil moisture dis tribution map was plotted with the Kriging method and then superposed over on the topographic factor maps of aspect, slopes and elevation, derived from DEM, thus forming a database of relationships between soil moisture and different topographic factors for analysis of effects of different topographic factors on soil moisture, and variat ion of soil moisture attributed to topographic factors.Results show that in the 0~30 cm soil layer, the topographic factors were in the order of aspect > elevat ion> slope in effect on soil moisture, while in the 30~60 cm soil layer, the sequence is slope> elevation> aspect The cluster analysis of aspects and slopes indicated that aspects could be sorted into two categories: 292.5°~360° (0°)~112.5°(shady slope) and 112.5°~292.5°(sunny slope), and so could slopes, 5°~20° and 20°~35° according to their effects on soil moisture.Based on the sorting models for fitting functional relations between soil moisture and the topographic factors were established and coefficients of the relations worked out.Assuming the soil moisture modulus is set as 1 when a shady slope is 5°~20° in gradient, and it is 0.99 on a sunny slope the same in gradient, 0.82 on a shady slope, 20°~35° in gradient, and 0.80 on a sunny slope the same in gradient.