Abstract: Struvite (MgNH₄PO₄·6H₂O) is commonly found in the recovered products after removal of nitrogen (N) and phosphorus (P) from wastewater by magnesium modified biochar, which can be used as a slow-release fertilizer. This study aims to investigate the effects of the struvite-based magnesium modified biochar (MAP-BC) combined with different fertilization rates on N and P reduction, rice yield and quality under different irrigation modes. MAP-BC was also applied in farmland under the different modes of irrigation and fertilization. Dongyan 18 (japonica rice) was used as the test material in the field experiment. Two irrigation modes were set, including the conventional flooding (I_CF) and alternate wet and dry (I_AWD) irrigation. Five types of fertilizer application were the conventional fertilization (N₁B₀), conventional fertilization + 5 t/hm² MAP-BC (N₁B₁), conventional fertilization + 10 t/hm² MAP-BC (N₁B₂), 25% reduction in N and P fertilizer + 5 t/hm² MAP-BC (N_3/4B₁), and 25% reduction in N and P fertilizer + 10 t/hm² MAP-BC (N_3/4B₂). Furthermore, MAP-BC was applied with a struvite purity of 82%. Excellent slow-release properties, N and P nutrient loads were obtained to solidify the subsequent field applications. Both MAP-BC and basal fertilizer were thoroughly mixed to simultaneously apply into the topsoil. Moreover, the MAP-BC was only applied in 2021 without 2022, in order to determine the positive effect of MAP-BC in the next growing season. The results showed that: Compared with I_CF, I_AWD significantly increased the chlorophyll content at milk-ripe stage, whereas, there was the significant decrease in the ineffective tiller number in 2021 (P<0.05). Compared with I_CF, I_AWD was significantly reduced the number of ineffective tillers by 17.39% in 2021 (P<0.05). MAP-BC was avoided the adverse effect of fertilizer reduction on chlorophyll content of rice, indicating the super compensation. Compared with N₁B₀, N_3/4B₁ and N_3/4B₂ treatments significantly increased the leaf chlorophyll content at tillering, jointing-booting, heading and flowering, and milk-ripe stage by 3.00%-6.83%, 2.99%-6.92%, 3.66%-5.69%, and 3.42%-9.02%, respectively (P<0.05). The high-purity struvite components in the MAP-BC were ensured the long-term supply of N and P nutrients to plants through the slow release. Compared with N₁B₀, N_3/4B₂ was fully met the demand for N and P nutrients at the later stage of rice growth. There was the significant increase in the N and P uptake at panicle by 4.77%-7.06% and 4.26%-12.69%, respectively (P<0.05). Compared with N₁B₀, the application of 10 t/hm² MAP-BC significantly increased the highest tiller number and the final tiller number at two years by 6.75%-9.64% and 13.16%-16.88%, respectively (P<0.05). In the two-year experiment, compared with N₁B₀ in I_AWD mode, the yield in the N₁B₁ and N₁B₂ treatments significantly increased by 7.66%-8.43% and 11.49%-12.64%, respectively (P<0.05). Meanwhile, 10 t/hm² MAP-BC was compensated for the adverse effects of 25% reduction in N and P fertilizer on yield. In I_AWD, N_3/4B₁ and N_3/4B₂ were significantly reduced the setback, chalky rice rate and chalkiness, whereas, there was the significant increase in the break down value (P<0.05). As such, the appearance and eating quality of rice were then improved significantly (P<0.05). In addition, N₁B₀ and N₁B₂ significantly increased(P<0.05) the protein content of rice by 2.66% and 5.79% compared with N₁B₀ treatment, indicating that the application of MAP-BC could improve the nutritional quality of rice. Therefore, 25% reduction of N and P fertilizer combined with 10 t/hm² MAP-BC in I_AWD was achieved in the fertilization reduction, yield increase and quality improvement under water-saving conditions. Thus, the finding can provide a theoretical basis for the green and efficient rice production. Subsequent studies should be considered on the utilization of actual eutrophic water in the MAP-BC preparation, in order to enhance the recycling efficiency of agricultural straw and eutrophic water.