Efficient irrigation management is critical for improving maize productivity in Rwanda's semi-arid regions, where rainfall variability frequently constrains crop yields. This study evaluated the effectiveness of irrigation scheduling strategies using the Soil Water Atmosphere Plant (SWAP) model in combination with the Soil Moisture Deficit Index (SMDI) at the Kagitumba Irrigation Scheme from 2016 to 2024. Weather, soil, and crop data were integrated to simulate soil moisture dynamics, maize yield, and irrigation water productivity (IWP) under five irrigation depths (7, 12, 20, 25, and 30 mm) and four irrigation intervals (3, 5, 7, and 9 days) using surface irrigation as the application method. Model calibration and validation demonstrated good performance (R-2 = 0.819; MRE = 3.74%), confirming its suitability for field-scale irrigation analysis. Results showed that the highest average yield (7164 kg ha(-)(1)) occurred under a 20 mm depth with a 3 days irrigation interval, while the highest average irrigation water productivity (10.36 kg ha(-)(1) mm(-)(1)) was achieved at 30 mm with a 5 days irrigation interval. These findings highlight the trade-off between yield maximization and water use efficiency. The study confirms the potential of SMDI-guided irrigation scheduling to minimize drought stress, optimize irrigation water productivity, and enhance maize yield by ensuring timely and precise water application that matches crop water requirements in Rwanda's drought-prone zones.