Many technologies have been developed to control agricultural non-point-source pollution (ANPSP). However, most reduce pollution from only a single source instead of considering an entire region with multiple pollution sources as a control unit. A pollutant reduction system for controlling ANPSP at a regional scale could be built by integrating technologies and the reuse of treated wastewater (TWR) and nutrients (NR) to protect the environment and achieve agricultural sustainability. The present study proposes four systematic schemes involving TWR for irrigation and NR in a region with three sources of ANPSP (crop farming, livestock and aquaculture). Subsequently, a multi-objective evaluation model is established based on the analytical hierarchy process (AHP) combined with grey relational analysis (GRA) to identify the optimal scheme considering six indices, namely, pollutant reductions (total nitrogen, TN; total phosphorous, TP; ammonium-nitrogen, NH₄ ⁺-N; and chemical oxygen demand, COD) and costs (construction and operational costs). The Taihu Lake Basin suffers from some of the worst ANPSP in China, and a case study was conducted in a town with three ANPSP sources. Four systems were developed on the basis of suggested technologies and the scenarios of TWR and NR (Scenario I: no reuse, Scenario II: reuse of all livestock wastewater and manure, Scenario III: reuse of some aquaculture wastewater, and Scenario IV: reuse of all livestock wastewater and manure and some aquaculture wastewater). Pollutant reductions were calculated based on removal efficiency and pollutant loads, which were estimated from the local pollutant export coefficients and agricultural information (crop farming, livestock, and aquaculture). The costs were determined on the basis of the total pollutant reductions and unit cost. The results showed that the optimal system was the Scenario IV because it had the highest grey correlation degree among the four proposed systems. The optimal system met the irrigation water demand in Xinjian. In the optimal system, the removal efficiencies of the pollutants TN, TP, NH₄ ⁺-N, and COD were 84.3%, 94.2%, 89.6% and 94.0%, respectively. In addition, the implementation of NR in the optimal system reduced the use of chemical fertilizers by nearly 81.7 kg N ha⁻¹ and 39.9 kg P ha⁻¹. The proposed methods provide a reference for the construction of a pollutant reduction system for controlling ANPSP in a multi-source region.