【Objective】Researches on moisture content and contraction characteristics of weathering crust of granite are the basis for evaluation of stability of granite soil and studies on erosion mechanism of the soil. However，direct measurement of the two parameters is very tedious, time consuming, and errors prone. Therefore, a number of indirect measuring methods have been developed to predict soil water retention curves by means of determining particle-size distribution, but most of these methods often yield intermittent soil moisture characteristic data because they are only capable of predicting soil moisture characteristic at individuals points along the curves. It is, therefore, essential to explore for a new method to plot characteristic curves of soil water content and contraction of weathering crust of granite. For that end, it is a first priority to understand relationships between soil moisture retention curve, contraction characteristic curve and particle-size distribution curve, as well as relationships between all the three curves and physical properties of the soil. 【Method】To comprehensively understand the relationships, soil samples different in weathering degree were collected from Southeast Hubei for analysis. Based on the similarity of the soil water retention curves, contraction characteristic curves and particle-size distribution curves, the Van Genuchten (VG) model was used to fit the curves of the soil samples different in weathering degree and explore fitting parameters of the curves and linear and non-linear correlations of the fitting parameters with soil physical properties. 【Result】 Results show that, the VG model may well fit all the three curves of granite soils different in weathering degree. Certain linear relationships were found between the fitting parameters, particularly between parameter a₁ of the particle-size distribution curves and parameter Q_S of soil water retention curves (r=-0.884, p<0.05) and but highly significant negative linear ones were between a₁ and parameter v_r of the contraction characteristic curves (r=-0.990, p<0.01); Meanwhile, certain linear relationships were also found between the curve fitting parameters and basic soil physical properties, especially those of the parameters of water retention curves with soil texture and bulk density. And the linear relationships of parameters v_r and v_sof the contraction characteristic curves with basic soil properties were closer than those of all the others; Besides non-linear relationships also exsited between fitting parameters of these curves and between the parameters and basic soil physical properties. Significant quadratic nonlinear relationships were found of parameter a₂ of the water retention curves with the clay and parametera₁ of the particle-size distribution curves, while significant cubic ones were between parameter a₃ of the contraction characteristic curves and sand, and between parameter v_s of the contraction characteristic curves and parameter n₁ of the particle-size distribution curves. 【Conclusion】 In this paper, the VG model was used to fit paricle-size distribution, soil moisture content and contraction characteritics of granite soils different in weathering degree and analyze linear and non-linear relationships between soil water retention curves, contraction characteristic curves and particle-size distribution curves, between fitting parameters of the three types of curves and between the fitting parameters and basic physical properties of the soil. All the findings will certainly be useful to future study on relationships between the three types of curves and have demonstrated that it is feasible to predict soil moisture content and contraction characteristics of granite soils based on their particle-size distribution and physical properties.