Aims and Background Plant nutrition models assume that phosphorus is depleted in the rhizosphere, due to root uptake. These models adequately predict P acquisition in P-rich soils but most often underestimate it in P- poor soils. We aimed at challenging the hypothesis of P depletion as the driving force of P acquisition in the rhizosphere. Methods The rhizosphere of various crops, either grown alone or intercropped, was sampled in rhizobox, pot and field experiments conducted in soils of contrasting P-status (the P-poor soil did not receive any fertilizer- P for 40 years, while the P-rich soil received excess P), in order to measure changes in P availability relative to bulk soil. Geochemical modelling w as used to understand the underlying processes. Results Phosphorus availability (water and Olsen extracts) almost systematically increased in the rhizosphere, for both the P-poor and P-rich soils. The absence of P depletion suggested that P bioavailability was ruled by another process than P uptake. Geochemical modelling helped eluc idating the contribution of root-induced pH changes and uptake of calcium. Conclusion The rhizosphere can be P-enriched rather than becoming P-impoverished, which contradicts common know ledge. Biogeoche mical processes other than P uptake have to be accounted for to adequately predict P nutrition in crops.