Understanding the molecular-scale heterogeneity of dissolved organic matter (DOM) in tea plantation soils is essential for explaining regional differences in tea quality and mitigating contaminant risks. However, systematic studies on the structural characteristics of DOM in tea garden soils at different molecular weight scales and its binding behavior with pollutants remain limited. In this study, the molecular weight fractionation, structural composition, and arsenic (As) binding behavior of DOM in iron-rich red soils from three premium tea-growing regions of Huangshan City, China, were comprehensively investigated. An integrated spectroscopic approach was employed to analyze soil samples from Huangshan Mao Feng (Region 1, HSMF), Keemun Black Tea (Region 2, QMHC) and Taiping Houkui (Region 3, TPHK) plantations. The results revealed that the soil DOM exhibited high aromaticity (average SUVA₂₅₄ > 4.34 L·mg^-¹·m^-1) and was predominantly of microbial origin (fluorescence index > 1.9). Distinct regional patterns emerged: Region 2 exhibited the highest aromaticity (SUVA₂₅₄ = 4.75 ± 3.99 L·mg^-¹·m^-1) and a pronounced microbial signature in low-molecular-weight DOM (LMW-DOM, FI = 1.89 ± 0.87). Region 1 was characterized by highly humified, high-molecular-weight DOM (HMW-DOM, HIX = 8.64 ± 3.57). In addition, Region 3 displayed stronger allochthonous inputs (S_R = 1.13 ± 0.58). PARAFAC analysis revealed that protein-like substances accounted for approximately 60 % of the fluorescence intensity in Region 2's HMW-DOM, whereas humic-like components dominated in Region 1 and Region 3. Furthermore, two-dimensional correlation spectroscopy revealed distinct DOM-As binding sequences: humic acid-dominated in Region 1 (389 nm → 324 nm → 288 nm), protein-dominated in Region 2 (425 nm → 390 nm → 386 nm) and fulvic acid-dominated in Region 3 (370 nm → 410 nm → 280 nm). These findings demonstrate that the molecular composition of DOM, which is shaped by local agronomic practices, determines its environmental reactivity. This provides crucial insights for optimising the management of tea quality and controlling the mobility of contaminants.