Municipal water systems in Rwanda are crucial for agriculture productivity and public health. However, their cost-effectiveness varies widely, necessitating a systematic evaluation. A randomized field trial was conducted in three districts. Water quality parameters were measured at five randomly selected farms per district over a six-month period. Economic data for each farm was collected and analysed using econometric models to assess cost-effectiveness. Water quality parameters showed consistent improvement with system upgrades, indicating effective implementation of the trial intervention. The randomized field trial demonstrated that upgrading municipal water systems can significantly improve agricultural productivity in Rwanda. Investment in infrastructure improvements and regular maintenance is recommended to sustain cost-effectiveness and ensure long-term benefits for farmers. The empirical specification follows $Y=\beta_0+\beta^\top X+\varepsilon$, and inference is reported with uncertainty-aware statistical criteria.