Although iron is the fourth element in the Earth’s crust, the availability of Fe(III) is limited in most cultivated soils due to their pH. In soil adhering to plant roots (rhizosphere), plants release a sgnificant part of their photosynthetates (rhizodeposits) that promote microbial density and activity. Iron uptake by this microflora and by plant roots contributes to decrease even more iron availability in the rhizosphere. In this competition context, plants and microbes have developed active strategies of iron uptake. In dicotyledon plants, this strategy involves (i) the excretion of protons, (ii) the reduction of Fe(III) by reductases, and (iii) plasmalemma transport of Fe(II) by iron transporters. For the microorganisms, this strategy is based on the synthesis of siderophores and on ferri-siderophore membrane receptors, in iron stress conditions. The major class of siderophores synthesized by fluorescent pseudomonads (pyoverdines) show a high affinity for Fe+++ leading to a reduced iron avaibility for other microorganisms such as pathogenic fungi. This efficient pyoverdine-mediated iron uptake was shown to be involded in the bacterial competitiveness and antagonism against pathogenic fungi. The corresponding demonstrations were made using pseudomonad mutants impaired in the siderophore synthesis and purified pyoverdines. Over the last years, we developped an innovative experimental approch relying on the variation of iron avaibility in the rhizosphere by the cultivation of transgenic plant overaccumulating iron (OV) – inducing an iron depletion in the rhizosphere - and of the corresponding wild-type (WT). The iron depletion induced by the OV line lead to a shift in the genetic structure of the pseudomonad community in the rhizosphere compared to the WT, as characterized by PCR-RFLP of DNA directly extracted from roots and amplified with specific primers. The impact of the OV line on the fluorescent pseudomonad community was further evaluated on a large collection of isolates. Populations preferentially associated with the OV line appeared to be less susceptible to iron stress and to produce specific pyoverdines expressing a significant higher iron-mediated antagonism than those from the WT, favouring therefore plant health. Since pyoverdines interfere with iron uptake of fungi, they were considered as being also likely to interfere with plant iron nutrition. This was evaluated with purified pyoverdine which appeared to not impair but even to improve the plant nutrition. Use of plant mutants showed that Fe from pyoverdine was not incorporated through the strategy described so far for dicotyledons. This was supported by the presence of pyoverdines in planta. Taken together our data indicate that in return of the cost for the plants represented by the rhizodeposits release, bacterial populations selected in the OV rhizosphere showed an efficient pyoverdine-mediated iron uptake that promoted plant health through fungal antagonism (iron competition) and plant growth through an improved plant iron nutrition (benefice). This feed-back lump is further investigated to quantify the costbenefice ratio for each partner (bacteria-plant) of this reciprocal interaction in the rhizosphere.