Legumes have been shown to increase P uptake of the following cereal, but the underlying mechanisms are unclear. The aim of this study was to compare the effect of legume pre-crops and their residues on the growth, P uptake and size of soil P pools in the rhizosphere of the following wheat. Three grain legumes (faba bean, chickpea and white lupin) were grown until maturity in loamy sand soil with low P availability to which 80 mg P kg−1 was supplied. This pre-crop soil was then amended with legume residues or left un-amended and planted with wheat. The growth, P uptake and concentrations of P pools in the rhizosphere of the following wheat were measured 6 weeks after sowing. In a separate experiment, residue decomposition was measured over 42 days by determining soil CO2 release as well as available N and P. Decomposition rates were highest for chickpea residues and lowest for wheat residues. P release was greatest from white lupin residues and N release was greatest from faba bean residues, while wheat residues resulted in net N and P immobilisation. The growth of the following wheat was greater in legume pre-crop soil without residue than in soils with residue addition, while the reverse was true for plant P concentration. Among the legumes, faba bean had the strongest effect on growth, P uptake and concentrations of the rhizosphere P pools of the following wheat. Regardless of the pre-crop and residue treatment, wheat depleted the less labile pools residual P as well as NaOH-Pi and Po, with a stronger depletion of the organic pool. We conclude that although P in the added residues may become available during decomposition, the presence of the residues in the soil had a negative effect on the growth of the following wheat. Further, pre-crops or their residues had little effect on the size of P pools in the rhizosphere of wheat.