This work is a quantitative evaluation of current hypotheses of iron acquisition by higher plants. The outputs of three root uptake models were compared to the iron uptake of a single plant species grown in three different soils. Iron hydroxide dissolution kinetics by organic ligands were formalised in the reactive transport modelling, together with ligand and ferric complex degradation. A global sensitivity analysis was performed. Iron uptake by Noccaea caerulescens does not appear to differ from that of other higher plants. The solubility of soil iron hydroxides alone cannot satisfy the plant iron requirements, even in an acidified rhizosphere. For this reason, all plants growing in aerated soils should suffer extreme iron deficiencies, not only those in alkaline soils. The dissociation rate of an Fe3+ complex (for instance with an aminopolycarboxilate ligand) in solution near the root cell membrane is not sufficient to provide the metal needed by the plant. In contrast, the adsorption of the ferric complex on the root cell membrane, before Fe3+ reduction and the absorption of Fe2+ appears to be a very efficient process. This is also the case for iron acquisition through the root secretion of an iron organic ligand dissolving soil iron hydroxides, the rate of ligand excretion being a key parameter. Mechanistic modelling supports the hypothesis of the reduction at the root cell membrane of a ferric complex resulting of iron hydroxide dissolution by a soil- or root-borne organic ligand.