Conventional oven-drying method may not give accurate information of bound water content in soils. The objective of this study was to determine the ftaction of soil water that is bound to soil colloids using the thetmogravimetry (TG) technique. A heating program from room temperature (RT) to 200°C was developed to partition bound water into loosely bound water (equivalent to water content determined by conventional oven-drying method) and tightly bound water that was not accounted by conventional oven-drying method on original and organic matter (OM) removed soil samples. Nine air-dry soils, with varying amounts of clay and OM contents, wete tested.For the original air-dry samples, bound water content ranged from 0.54 to 5.22%, and the fractions of loosely and tightly bound water were about 80 and 20%, respectively. On a mass fraction basis, the specific watet adsorption capacity o( soil OM was 10 to 40 times that of soil minerals. For most mineral soils, however, the contribution of soil minerals to bound water (>70%) was much larger than that of soil OM (<30%), since the mineral fraction usually dominates over the OM fraction. In a soil with high OM concentration but a relatively low clay content.OM contribution to bound water exceeded 50%. Soil specific surface area (5/i) showed a strong infiuence on bound water content, and a linear relationship between SÀ and bound water content was established.Abbreviations: DSC, difFerential scanning calorimetry; DTA, differential thermal analysis; EGD, evolved gas detection; MS, mass spectrometry; OM, organic matter; RT, room temperature; SA. specific surface area; SAR, specific adsorption ratio; TG, thermogravimetry.T he structure, energy, and properties (e.g., specific heat capacity and dielectric permittivity) of soil water vary with the distance from solid surface (Etzler, 1983; Yamaguchi, 1959). Water bounded in the first few layers is assumed to have stronger bonding energy (Panchev et al., 2005) and therefore different structure than the water in bulk (Bogdan et al., 1996;Kleinberg and Griffin, 2005). Soil water content is usually referred to the quantity of water that is evaporated from soil by heating the sample at 100 to 110°C (average 105°C) until there is no further weight loss. Some research has implied that the drying temperature of porous medium perhaps should be higher than 105°C. For example, thermo-analytical studies of water on activated alumina showed that the drying temperature for this material should be 120 to 125°C (Hampson and Bleam, 1996). Bogdan et al. (1996) measured water content of pyrogenic silica powder by drying at 200°C. Gardner et al. (2001) pointed out that heating at temperature range of 110 to 16O°C was required to remove bound water from clay surfaces. Logsdon and Laird (2004) dried clay samples at 190°C. Mitchell and Soga (2005) stated that depending on soil texture, the drying temperature for mineral soils should be in the range of 100 to 300°C. Thus it is questionable if the conventional method of oven-drying soil samples at 105...