Both batch and column methods are conventionally utilized to determine some critical parameters for assessing the transport of contaminants of concern. The validity of using these parameters is somewhat confusing, however, since outputs such as distribution coefficient (K-d) from these two approaches are often discrepant. To bridge this gap, all possible factors that might contribute to this discrepancy were thoroughly investigated in this report by a case Study of Cs sorption to crushed granite Linder various conditions. Our results confirm an important finding that solid/liquid (S/L) ratio is the dominant factor responsible for this discrepancy. As long as the S/L. ratio exceeds 0.25. a consistent K-d value can be reached by the two methods. Under these conditions (S/L ratios > 0.25), the sorption capacity of the solid is about an order of magnitude less than that in low S/L ratios (<0.25). Although low sorption capacity is observed in the cases of high S/L ratios. the sorption Usually takes place preferentially on the-most favorable (thermodynamically stable) sorption sites to form a stronger binding. This is verified by our desorption experiments in which a linear isotherm feature is shown either in deionized water or in 1 M of ammonium acetate Solutions. It may be concluded that batch experiment with an S/L ratio exceeding 0.25 is crucial to obtain convincing Kd values for safety assessment of radioactive waste repository. (C) 2008 Elsevier B.V. All rights reserved.