Tropical grasses grown as cover crops can mobilize phosphorus (P) in soil and have been suggested as a tool to increase soil P cycling and bioavailability. The objective of this study was to evaluate the effect of tropical grasses on soil P dynamics, lability, desorption kinetics and bioavailability to soya bean, specifically to test the hypothesis that introducing grass species in the cropping system may affect soil P availability and soya bean development according to soil P concentration. Three grass species, ruzi grass (Urochloa ruziziensis), palisade grass (Urochloa brizantha) and Guinea grass (Megathyrsus maximus), were grown in soils with contrasting P status. Soya bean was grown after grasses to assess soil P bioavailability. Hedley P fractionation, microbial biomass P, phytase-labile P and the diffusive gradient in thin films were determined, before and after cultivation. It was found that grasses remobilized soil P, reducing the concentration of recalcitrant P forms. The effect of grasses on changing the P desorption kinetics parameters did not directly explain the observed variation on P bioavailability to soya bean. Grasses and microorganisms solubilize recalcitrant organic P (Po) forms and tropical grasses grown as cover crops increased P bioavailability to soya bean mainly due to the supply of P by decomposition of grass residues in low-P soil. However, no clear advantages in soya bean P nutrition were observed when in rotation with these grasses in high-P soil. This study indicates that further advantages in soya bean P nutrition after tropical grasses may be impeded by phytate, which is not readily available to plants. © 2018 British Society of Soil Science