A test of fundamental soil chemical assumptions used in the model by Reuss and Johnson, the Birkenes model, MAGIC and ILWAS has been carried out on leachate data from soil lysimeters installed at Nordmoen, southeast Norway. Simple cation exchange theory was found only to explain the relations between the concentrations of Ca2+ and Mg2+ in leachate. Assumptions of Al control through gibbsite equilibrium were, as in other recent studies, shown to fail in this study as well. Multiple linear regression was alternatively used in order to find relations between variation in soil and leachate chemistry. The models proposed for each sampling depth were so different, however, that no general model for the leaching of various chemical elements could be developed. This study emphasized the importance of intensive studies of water flow paths and retention time, spatial soil variability, changes in redox potential, and processes like decomposition and mineralization, in order to improve the mathematical models related to the acidification of soil-water systems.