A 2-year pot experiment was conducted to study the differences and relationship of inorganic phosphorus forms, available phosphorus in rhizosphere soil and phosphorus uptake in maize and soybean intercropping system under different phosphorus levels, in which 9 treatments, three cropping patterns (maize monocropping, soybean monocropping and maize-soybean intercropping)×three P₂O₅ levels (0, 50 and 100 mg/kg, labeled P0, P50 and P100). The results showed that compared with monocropping, intercropping significantly increased grain yields and phosphorus uptake of maize and soybean under P0, P50 and P100 levels in 2018 and 2019. Phosphorus uptake and grain yields of intercropped maize and soybean were not declined in the P50 level compared to corresponding monocropped plants in P100 level. Intercropping increased available phosphorus content in rhizosphere soil of maize and soybean under P0, P50 and P100 levels, but decreased the contents of total inorganic phosphorus, Fe-P, Al-P, Ca-P and O-P. Meanwhile, appropriate application of phosphorus fertilizer significantly increased the contents of total inorganic phosphorus and inorganic phosphorus forms in rhizosphere soil of maize and soybean. Under this experimental condition, the activation of Fe-P, Al-P, Ca-P and O-P in soil (especially Fe-P) induced by intercropping was an important source for the increase of soil available phosphorus content and phosphorus uptake in intercropping under low phosphorus condition. Maize-soybean intercropping has the potential to save phosphorus fertilizer while maintain crop yields, rhizosphere available phosphorus and phosphorus uptake.