While most studies have tried to assess phosphorus bioavailability based on phosphorus availability determined in the bulk soil, we question this approach which does not account for changes of P availability that occur in the rhizosphere. In this study, we combined the extraction of soil inorganic phosphorus (P) with CaCl2 or water and geochemical modelling in order to unravel the variations of P availability in the bulk soil and rhizosphere of durum wheat (Triticum turgidum durum L.) over a range of soil pH (pH ~ 4 to 9) chemically modified from initial pH. For this purpose, mechanistic models were used to simulate the adsorption of cations and anions by soil constituents with an additive approach. The geochemical codes accurately reproduced the concentration of extracted P by CaCl2 and water for both bulk soil and rhizosphere (RMSE = 5.3 and 10.6 μg.kg-1 for bulk and rhizosphere soil respectively). In rhizosphere, the effect of pH on P availability obtained by both extraction methods exhibited two different trends, which contrasted with that on P bioavailability. Phosphorus uptake by durum wheat mainly explained the difference of P availability between the rhizosphere and bulk soil for CaCl2 extraction. For water extraction, Ca uptake as well as P uptake need to be accounted to explain the observed difference of P availability between rhizosphere and bulk soil. Our results highlight the extra benefits of mechanistic models to clearly understand how P availability can change in the rhizosphere, in order to better predict P bioavailability to crops.