Ecosystem services (ES) of rice-wheat rotation agroecosystems play an important role in global ecosystem services. Some ES are facing large-scale degradation owing to changes in land use and agriculture management modes in the Taihu Lake Basin. This has resulted in an increasing awareness of sustainable agricultural development through optimizing and advancing ES of the agroecosystems. This study simulated the trends and trade-offs of agroecosystem ES in the Taihu Lake Basin from 1986 to 2015 using the APSIM model. The APSIM model was calibrated by PEST software, with observations on grain supply (grain field), climate regulations (N₂O emissions, carbon sequestration-soil organic carbon), and water quality regulations (N-losses). Recently, high utilization of chemical fertilizers and low-use efficiency of straw has led to many severe issues in the Taihu Lake Basin, such as agricultural non-point source pollution. Based on the combinations of differential straw-returns (non-straw returning, rice straw-returning and non-wheat straw-returning, wheat straw-returning and non-rice straw-returning, all straw-returning) and different fertilizer use (0, -5%, -10%, -20%), this study designed 16 scenarios to simulate and assess three key ES in the Taihu Lake Basin. The trade-offs were assessed by using a self-organizing map (SOM) to derive effective management measures to improve overall agroecosystem ES. The results show that:(1) grain yield, N₂O emissions, N losses, and soil organic carbon are more significantly affected by fertilizer use than are straw-returns, and the former are characterized by declines with reduced fertilizer utilization; (2) Different management modes have a weak impact on carbon sequestration, but significant effects on N₂O emissions and water quality regulation. All straw-returning and 5%, 10%, or 20% reductions of fertilizer utilization are more effective in suppressing N₂O emissions, and the reduction of N losses is greater than 7%. (3) There exists a strong synergistic relationship between grain supply and both carbon sequestration and N₂O emissions, respectively. Grain supply and water quality regulation, carbon sequestration and water quality regulation, and water quality regulation and N₂O emissions present significant trade-offs, respectively. Carbon sequestration and N₂O emissions show a weak trade-off. (4) Assuming water quality regulation (N losses) and climate regulations (N₂O emissions) are lower than those of the basic scenario and the reduction in grain yield is less than 5%, two scenarios-non-straw returning and 5% fertilizer reduction (S0_F1), and all straw-returning and 5% fertilizer reduction (S3_F1)-can help to advance overall agroecosystem services in the Taihu Lake Basin. It is necessary to improve calibration of the APSIM model because of the relatively low soil organic carbon. Additionally, the comprehensive effects of different modes on agroecosystem ES in the Taihu Lake Basin need to be further explored.