Inorganic P is the least mobile major nutrient in most soils and is frequently the prime limiting factor for plant growth in terrestrial ecosystems. In this study, the extraction of soil inorganic P with CaCl2 (P-CaCl2) and geochemical modelling were combined in order to unravel the processes controlling the environmentally available P (EAP) of a soil over a range of pH values (pH similar to 4-10). Mechanistic descriptions of the adsorption of cations and anions by the soil constituents were used (1-pK Triple Plane, ion-exchange and NICA-Donnan models). These models are implemented into the geochemical code Visual MINTEQ. An additive approach was used for their application to the surface horizon of a Cambisol. The geochemical code accurately reproduced the concentration of extracted P at the different soil pH values (R-2 = 0.9. RMSE = 0.03 mg kg(-1)). Model parameters were either directly found in the literature or estimated by fitting published experimental results in single mineral systems. The strong agreement between measurements and modelling results demonstrated that adsorption processes exerted a major control on the EAP of the soil over a large range of pH values. An influence of the precipitation of P-containing mineral is discounted based on thermodynamic calculations. Modelling results indicated that the variations in P-CaCl2 with soil pH were controlled by the deprotonation/protonation of the surface hydroxyl groups, the distribution of P surface complexes, and the adsorption of Ca and Cl from the electrolyte background. Iron-oxides and gibbsite were found to be the major P-adsorbing soil constituents at acidic and alkaline pHs, whereas P was mainly adsorbed by clay minerals at intermediate pH values. This study demonstrates the efficacy of geochemical modelling to understand soil processes, and the applicability of mechanistic adsorption models to a 'real' soil, with its mineralogical complexity and the additional contribution of soil organic matter