Sorption coefficients are used to describe the equilibrium distribution of a chemical between a soil or sediment and the aqueous phase that it is in contact with. Although sorption coefficients for a particular organic chemical vary greatly from soil to soil, the observation has been made that sorption generally increases as the organic carbon content of the soil and the hydrophobicity of the chemical increases. This general observation resulted in the acceptance of organic carbon normalized sorption coefficients (Koc) as unique properties or constants of organic chemicals. In turn, Koc values have been estimated by quantitative structure-activity relationships (QSARs) developed by correlation with a variety of physical or chemical properties and structural descriptors related to the hydrophobicity of the chemical such as octanol-water partition coefficients, aqueous solubilities, molecular connectivity indices, molecular weight, molecular surface area, and reverse-phase high-performance liquid chromatography retention times. The selection and application of the most appropriate QSAR for predicting Koc depend on several factors, including the availability of required input, the appropriateness of model to chemical of interest, and the methodology for calculating the necessary topological or structural information. A review of the existing QSARs for predicting Koc and the limitations of using the Koc approach to estimate sorption coefficients will be presented.