Abstract: This study employed a long-term positioning experiment using dryland, low-phosphorus red soils under varying planting configurations (maize monoculture and maize–soybean intercropping) and distinct phosphorus application levels (superphosphate at P₂O₅ 0, 60, and 90 mg/kg²) as the experimental subjects. Through pot experiments inoculated with field soil, the effects of two phosphorus supply levels (0 and 40 mg/kg, referred to as P₀ and P₆₀, respectively) on maize aboveground biomass, phosphorus uptake, partial land equivalent ratio, mycorrhizal colonization, soil available phosphorus, and phosphatase activities were analyzed, and the soil legacy effects formed by continuous planting of different planting patterns over the medium to long term in the farmland ecosystem of monoculture and intercropping systems were investigated. The results showed that, compared with the sterilized treatment, inoculation with native soil significantly increased maize aboveground biomass, phosphorus uptake, mycorrhizal colonization, and soil available phosphorus by 203.05%, 128%, 124.30%, and 21.05%, respectively, and increased the partial land equivalent ratio by 0.21. Inoculation with native soil from the P₆₀ maize–soybean intercropping treatment significantly enhanced both acid and alkaline phosphatase activities compared to the other treatments. Under native soil inoculation, compared with monoculture, intercropped maize aboveground biomass, phosphorus uptake, mycorrhizal colonization, soil available phosphorus, and acid and alkaline phosphatase activities increased by 14.17%, 104.49%, 17.55%, 11.52%, 5.80%, and 38.14%, respectively. In conclusion, both phosphorus supply levels and planting patterns jointly influenced the soil legacy effect, indicating that reasonable intercropping combined with moderate fertilization can effectively improve plant–soil relationships, thereby promoting plant growth, nutrient uptake, and activation.