Abstract: Cropping patterns and fertilizer management are critical for soil phosphorus (P) use efficiency. In order to explore the dynamic changes of inorganic P fractions in crop rhizosphere soils and to establish theoretical basis for optimal P management in maize/soybean relay intercropping, we designed a pot experiment where aboveground biomass and P uptake of maize and soybean, and available P and inorganic P fraction concentrations in rhizosphere and non-rhizosphere soils were determined. The experiment included three cropping patterns of maize/soybean relay intercropping (M/S), maize monoculture (MM) and soybean monoculture (SS). There were also three fertilization treatments, including non-fertilization (CK), nitrogen and potassium fertilization (NK) and then nitrogen, phosphorus and potassium fertilization (NPK). The results showed that grain yield of intercropped maize was significantly higher than that of mono-cropped maize under the same fertilization treatment. P fertilization significantly increased grain yield of mono-cropped maize whereas it had no significant effect on grain yield of intercropped maize. Stalk biomass and grain yield of intercropped soybean were all higher than those of mono-cropped soybean, regardless of fertilization application. Shoot P accumulation of intercropped maize and soybean per plant was significantly higher than that of mono-cropped system under all fertilization treatments. At maturity stage of maize, rhizosphere soil available P concentrations under intercropped maize were respectively 54.2% and 71.8% higher than those of mono-cropped maize under CK and NK treatments. At the start of blooming stage of soybean, rhizosphere soil available P concentration under intercropped soybean was 19.8% higher than that of mono-cropped soybean under NPK treatment. At maturity stage of soybean, rhizosphere soil available P concentrations of intercropped soybean were respectively 23.8% and 108.0% higher than that of mono-cropped soybean under NK and NPK treatments. Al-P concentrations in rhizosphere soils were lower than those of non-rhizosphere soils under both intercropped and monocultured maize for all the three fertilization treatments. Al-P concentrations in rhizosphere soils under monoculture maize were respectively 1.19 and 1.22 times those under intercropped maize for both CK and NK treatments. Fe-P concentration in rhizosphere soils under monoculture maize was 1.21 times that under intercropped maize for NPK treatment. Under the CK, NK and NPK fertilization treatments, Al-P concentrations in rhizosphere soils under monoculture soybean were respectively 1.12, 1.30 and 1.25 times those under intercropped soybean. Also Al-P concentrations in non-rhizosphere soils of monoculture soybean were respectively 1.22, 1.30 and 1.06 times those under intercropped soybean for CK, NK and NPK treatments. Similarly, Fe-P concentrations in rhizosphere soils under monoculture soybean were respectively 1.47 and 1.12 times those under intercropped soybean for CK and NK treatments. In conclusion, maize/soybean relay intercropping was favorable for soil Al-P and Fe-P dissolution and uptake under low P dose, compared with monoculture systems.