The quantitative contribution of the forest floor to P nutrition of maritime pine seedlings was experimentally determined by Jonard et al. (2009) in a greenhouse experiment using the radio-isotopic labeling. To extend the results of the experiment on a known mineral soil, a modeling approach was developed to predict P uptake of maritime pine seedlings growing in a mineral soil covered with a forest floor layer. The classical nutrient uptake model based on the diffusion/mass-flow theory was extended to take into account mineralization of P in dead organic matter, microbial P immobilization and re-mineralization and P leaching. In addition, the buffer power characterizing the P retention properties of the mineral soil was allowed to vary with time and with the P-ion concentration in solution. To account for increasing root competition with time, a moving boundary approach was implemented. According to the model, the forest floor contributed most of the P supply to the seedlings (99.3% after 130 days). Predicted P uptake was consistent with observed P uptake and modeling efficiency was 0.97. The uptake model was then used to evaluate the impact of the P retention properties of the mineral soil on the contribution of the forest floor to P uptake. Simulations showed that the contribution of the forest floor was much lower in the quasi non-reactive soil (45.7%) but rapidly increased with soil P reactivity.