Coastal beach pollution caused by primordial radionuclides has emerged as a global environmental concern. This study presents the first integrated probabilistic–spatial framework for assessing radiological risk in the Bay of Bengal (BoB) beach sands. Twenty-two surface sediment samples (10–15 cm depth) from Darianagar, Himchari, and Nidania beaches along the BoB were analyzed using HPGe gamma spectrometry. Geometric mean (range) activity concentrations (Bq kg−1) were 77.6 (15.1–1914.3) for 226Ra, 123.1 (8.12–1847.3) for 232Th, and 312.8 (122.4–562) for 40K, exceeding global averages for 226Ra and 232Th by factors of 2.2 and 4.1 times, respectively. Principal component analysis (PCA) identified natural mineralogical controls, primarily monazite and K-feldspar rather than distinct anthropogenic sources, while Monte Carlo simulations revealed 95th-percentile radiological indices (Raeq, D_out, ELCR) exceeding international safety thresholds (8.5 × 10−4 to 23.3 × 10−4). Deterministic calculations yielded mean values higher than probabilistic medians, reflecting right-skewness. All radionuclide concentrations followed lognormal distributions, indicating significant spatial variability. Sensitivity analysis showed 232Th (∼75 %) as the primary contributors to total radiological risk. GIS-based spatial modeling delineated localized high-risk hotspots requiring restricted access, hazard signage, and periodic radiological monitoring. This study establishes benchmark data and a transferable methodological framework for coastal radiation surveillance and sustainable management, supporting evidence-based regulation of BoB and similar global beach systems.