Study the variation characteristics of soil phosphorus (P) forms are crucial for measuring soil quality and P absorption. However, studies on the effects of rice–duck co-culture system on soil P forms in flooded and dryer period as well as on the relationships among P forms, P absorption and environmental factors remain insufficient. A rice field experiment with consistent P inputs in each fertilizer treatment: no fertilizer (RD and RM), chemical fertilizer (RDF and RMF), organic fertilizer (RDO and RMO), and a mixture of 70% chemical and 30% organic fertilizers (RDFO and RMFO) were conducted to study it. The results showed that RDO and RDFO had higher readily available inorganic P pool (Resin P + NaHCO 3-P i), extractable organic P fractions (NaHCO 3-P o + NaOH-P o), and Fe-P and Al-P (NaOH-P i) than the corresponding rice monoculture. RDF decreased the extractable organic P fractions (NaHCO 3-P o + NaOH-P o) and Fe-P and Al-P (NaOH-P i) compared with RMF; RDO increased the accumulation of Residual-P while RDF and RDFO decreased the accumulation of Residual-P compared with RMO, RMF, and RMFO. pH was the dominant factor affecting P forms of soil. RDF and RDFO had higher P content of grain and straw that mainly affected by soil NaHCO 3-P i and NaOH-P i than other treatments. Simultaneously, rice–duck co-culture system can increase phosphorus use efficiency (PUE), especially RDF significantly increased by 171.63% and 22.63% than RDO and RDFO, respectively. In conclusion, our results suggested that rice–duck co-culture can promote rice grain and straw to absorb soil P by changing the soil pH to stimulate soil surplus P into active P. RDFO was the most recommended agricultural production model for comprehensive production.