The aim of this paper is to assess the suitability of DGT to extract kinetic rates of desorption of cesium (Cs) from soils. For this purpose, laboratory experiments with a natural soil spiked with Cs were carried out under three different contamination conditions, reflecting either an increase in Cs contamination level or an ageing of the contamination within the soil. The experimental results (i.e. the Cs accumulation kinetics onto DGT probes) were interpreted by the DGT-PROFS model. The latter calculates the partitioning of Cs between two particulate pools, describing weak and strong interactions respectively, as well as kinetic rates describing exchange reactions. It was shown that experimental conditions did not show any major impact on desorption rates, suggesting that desorption kinetics were not significantly affected by contamination level and ageing. Instead, the distribution of Cs among weak and strong sites was shown to be the predominant factor governing the differences observed in the remobilization of Cs to porewater among experimental conditions. The DGT technique combined with the DGT-PROFS modelling approach was proved to be efficient in estimating the desorption kinetic rates of Cs in soils.