Lupinus albus L. were grown in rhizoboxes containing a soil amended with sparingly available Fe-P or Al-P (100 μg P g −1 soil/resin mixture). Root halves of individual plants were supplied with nutrient solution (minus P) buffered at either pH 5.5 or 7.5, to assess whether the source of mineral-bound P and/or pH influence cluster-root growth and carboxylate exudation. The P-amended soil was mixed 3:1 (w/w) with anion-exchange resins to allow rapid fixation of carboxylates. Treatments lasted 10 weeks. Forty percent and 30% of the root mass developed as cluster roots in plants grown on Fe-P and Al-P respectively, but cluster-root growth was the same on root-halves grown at pH 5.5 or 7.5. Mineralbound P source (Al-or Fe-P) had no influence on the types of carboxylates measured in soil associated with cluster roots-citrate (and trace amounts of malate and fumarate) was the only major carboxylate detected. The [citrate] in the rhizosphere of cluster roots decreased with increased shoot P status (suggesting a systemic effect) and also, only for plants grown on Al-P, with decreased pH in the root environment (suggesting a local effect). In a separate experiment using anion exchange resins pre-loaded with malate or citrate, we measured malate (50%) and citrate (79%) recovery after 30 days in soil. We therefore, also conclude that measurements of [citrate] and [malate] at the root surface may be underestimated and would be greater than the 40-and 1.6-μmol g −1 root DM, respectively estimated by us and others because of decomposition of carboxylates around roots prior to sampling.