Human activities and climate change are acting as factors that affect soil evolution (pedogenesis). Their consequences are however poorly known and quantified. To be able to predict soil evolution in the future as a function of different climate scenarios and systems of land use management, modelling approaches are necessary. Different types of models can be used. The emphasis of the workshop was on deterministic and/or mechanistic models, distinguishing long (1000-10000 years) and short (10-100 years) time scales into two sessions (oral presentations, posters and discussions). Session 1: Modelling pedogenesis on long time scale For the long time scale, some models exist that for example focus on the understanding of major processes of soil formation resulting in actual soil three-dimensional spatial distributions through mass balance approaches. Session 2: Modelling pedogenesis on short time scale For the short time scale, various models that are mainly used in environmental considerations could be used to predict recent pedogenesis. These approaches consist for example of water flow models coupled with particle and solute transport. The use of geochemical models, like PhreeQc, MINTEQ, etc., allows kinetic approaches to soil processes to be used. Indeed in most temperate soils, equilibrium rarely occurs. This workshop allowed discussions of the advantages and limits of the different approaches that have been compared only rarely with real cases. This symposium gathered scientists coming from different disciplines including geology, hydrology, water chemistry and soil science.