To investigate the impact of long-term conservation tillage on spring wheat water productivity and evaluate yield stability under different precipitation patterns, this study utilised a long-term conservation tillage plot trial established in 2001 at the Dingxi Dryland Agriculture Comprehensive Experimental Station of Gansu Agricultural University. Four treatments were implemented: conventional tillage (T), no-tillage (NT), conventional tillage with straw incorporation (TS), and no-tillage with straw mulching (NTS). Field data from 2024-2025 were used to calibrate and validate the APSIM model, which was then applied to simulate soil organic carbon and spring wheat water productivity under different tillage practices from 2002 to 2025. The results showed that the APSIM model demonstrated good accuracy and adaptability in the study area, with the normalized root mean square error (NRMSE) between observed and simulated phenological phase values of spring wheatt in 2024-2025 below 15%. Simulated spring wheat yield under different tillage practices showed significantly positive correlations with observed values over the two years, with coefficients of determination of 0.899 and 0.978, respectively. The NRMSE was 4.6% and 2.3%, respectively, and the coefficient of determination exceeded 0.90 in all cases. The NRMSE between model-simulated and observed soil water storage throughout the spring wheat growth period was 6.5% and 7.2%, respectively, indicating the model's capability to accurately simulate indicators related to soil moisture productivity in wheat fields. Long-term simulation results indicated precipitation was the key factor influencing spring wheat production in this region. Farmland productivity stability and sustainability peaked in normal-precipitation years and were lowest in drought years. Spring wheat yield showed no significant difference between wet and normal years but was markedly higher than that in drought years, with an increase rate of 23.2% and 22.4%, respectively. The no-tillage straw mulching (NTS) treatment demonstrated significant advantages under various tillage practices and precipitation patterns. Specifically, it exhibited higher organic carbon storage (0-30 cm), yield, and water use efficiency than other treatments. Moreover, it demonstrated superior yield stability and field sustainability when confronting water shortages. Over all no-tillage straw mulching is an effective tillage practice suitable for achieving high and stable crop yield in the local context.