Aims: This study aimed to investigate the effects of coexistence with faba bean, a phosphorus (P)-efficient crop, on soil-accumulated P use by a maize/faba bean intercropping system on dynamic changes in soil P pool. Methods: Maize and faba bean were grown in P-accumulated soil as either sole cropping or intercropping. After one year (Stage I) or four years (Stage II) of no P application, soil samples were collected respectively and analyzed for soil P pools using sequential fractionation. Aboveground biomass and P content were annually measured from 2013 to 2016 to assess the annual P balance. Results: The intercropped maize/faba bean system showed a P-uptake advantage, with a Land Equivalent Ratio (LER) ranging from 1.2 to 1.5. The average shoot P content over the four years in intercropped maize and faba bean was significantly greater than that of the corresponding sole crops by 29% and 30%, respectively. Over the three-year P depletion period, the three cropping systems primarily depleted the 1 M HCl-Pi fraction, followed by sole maize, which depleted the NaOH-Pi and concentrated HCl-Po fractions. Sole faba bean depleted the alkali-soluble Po fraction (extracted by NaHCO3 and NaOH), and the intercropped maize/faba bean system depleted the conc. HCl-Po fraction, which was similar to the effect of sole maize. Conclusions: Both sole crops and intercrops mainly depleted 1 M HCl-Pi, but differed in Po depletion. Sole maize and maize/faba bean intercropping depleted the sparingly labile Po fraction, while sole faba bean depleted the labile and moderately labile Po fractions.