Aims : To test if multi–surface models can provide a soil-specific prediction of metal mobilization by phytosiderophores (PS) based on the characteristics of individual soils. ; Methods : Mechanistic multi-surface chemical equilibrium modeling was applied for obtaining soil-specific predictions of metal and PS speciation upon interaction of the PS 2’-deoxymugineic acid (DMA) with 6 soils differing in availability of Fe and other metals. Results from multi-surface modeling were compared with empirical data from soil interaction experiments. ; Results : For soils in which equilibrium was reached during the interaction experiment, multi-surface models could well predict PS equilibrium speciation. However, in uncontaminated calcareous soils, equilibrium was not reached within a week, and experimental and modeled DMA speciation differed considerably. In soils with circum-neutral pH, on which Fe deficiency is likely to occur, no substantial Fe mobilization by DMA was predicted. However, in all but the contaminated soils, Fe mobilization by DMA was observed experimentally. Cu and Ni were the quantitatively most important metals competing with Fe for complexation and mobilization by DMA. ; Conclusion : Thermodynamics are unable to explain the role of PS as Fe carrier in calcareous soils, and the kinetic aspects of metal mobilization by PS need to be closer examined in order to understand the mechanisms underlying strategy II Fe acquisition.