<p>Accurate prediction of groundwater nitrate concentrations is important for safe drinking of groundwater in the future and effective safeguarding of regional water resources. A random forest regression model was used to model the nitrate concentrations in shallow groundwater in the Hotan area of the Xinjiang Uygur Autonomous Region, China, in 2014, 2017, and 2022 in this study. The impact of land use on groundwater nitrate concentrations was quantified through contributing zones, and a Bayesian optimization algorithm was used to optimize the hyperparameters. The input parameters include various aspects such as topography, climate, soil, land use, etc., and the output element is the nitrate concentration. The importance ranking of each input element was obtained, and a predicted distribution map of NO₃⁻ concentration was generated based on the results. The performance of the model was assessed using the mean absolute error, root mean square error, and coefficient of determination R². R² &gt; 0.95 was obtained for the three years, and better model predictions were obtained. The elevation, subsoil organic carbon content, precipitation, topsoil bulk weight, and pH were the most important factors affecting the groundwater nitrate concentration. The proportion of bare land among the land use types was also a more important factor influencing the nitrate concentration, accounting for 5.7338%, as it occupied an extremely large portion of the study area. This contribution was crucial for the prediction of the model. Nitrogen fertilizer use should be controlled and ecological buffer zones should be established for areas with high nitrate pollution in the study area.</p>