Mutations in the arylsulfatase B (ARSB) gene are directly implicated in mucopolysaccharidosis type VI (MPS VI). Several non-synonymous single-nucleotide polymorphisms (nsSNPs) in ARSB have been associated with disease pathogenesis. A comprehensive evaluation of these variants is essential to understand their structural and functional consequences. In this study, a systematic in silico analysis was performed to identify deleterious nsSNPs in the ARSB gene. Initially, 430 nsSNPs were evaluated using sequence-based prediction tools, including SIFT, PolyPhen-2, FATHMM, and Mutation Assessor. Subsequently, 141 nsSNPs were subjected to structure-based stability analysis using MAESTROweb, SDM2, mCSM, and DynaMut2, of which 57 variants overlapped with previous reports. High-confidence deleterious nsSNPs were further assessed for pathogenicity using PMut and MutPred2 servers. Our integrated computational approach identified 44 highly deleterious mutations. Aggregation propensity analysis revealed that 29 of these variants exhibit increased aggregation tendencies, while one variant demonstrated progressive loss of solubility. Molecular dynamics simulations further indicated that high-confidence deleterious nsSNPs significantly disrupt ARSB structural integrity, enhance molecular flexibility, reduce structural rigidity, and promote atomic-level aggregation. Overall, this study provides mechanistic insights into how pathogenic mutations destabilize the ARSB protein and contribute to MPS VI pathogenesis, highlighting potential targets for future therapeutic investigation.