<p>The present study focuses on analyzing groundwater dynamics in the Eastern Duars region of West Bengal, India, from 2003 to 2023. The study, at first, performed a trend analysis of groundwater storage using either the Mann–Kendall (MK) or Modified Mann–Kendall (MMK) tests, depending on the results from autocorrelation function (ACF) plots and the Ljung-Box test. Thereafter, analyses included groundwater storage anomalies for gain–loss patterns, Pearson correlation to assess relationships with influencing factors, and an integrated Minimum Redundancy Maximum Relevance (mRMR)-based Multiple Linear Regression (MLR) approach to evaluate the combined impact of independent variables on groundwater. Along with the above test, to understand local perceptions of water scarcity and the drying up of wells, a structured field survey was conducted across locations within the study area. The results revealed significant groundwater depletion from January to May and in December, with the pre-monsoon and winter seasons showing the steepest declines. Monsoon trends were insignificantly negative, while post-monsoon trends were mixed. Annually, all trends were negative, indicating persistent depletion. Change point analysis and groundwater anomaly graphs revealed most shifts around 2009/2010, signalling recent declines. Profile and root zone soil moisture had the strongest impact on groundwater storage (correlations: 0.957–0.998). Plant canopy surface water and base flow-groundwater runoff showed strong to moderate correlations (0.710–0.737 and 0.663–0.703), while rainfall had moderate correlations (0.313–0.399). Cumulative rainfall over three periods (CR3) showed stronger correlations (0.666–0.761), emphasizing the role of accumulated rainfall on groundwater storage. These findings support sustainable groundwater management strategies.</p>