This study aimed to investigate the association between XRCC1 Arg399Gln and Arg194Trp single nucleotide polymorphisms (SNPs) and the risk and severity of polyneuropathy (DPN) in patients with type 2 diabetes mellitus (T2DM). The genotyping of SNPs was achieved in 732 contributors, including diabetic subjects with and without polyneuropathy and controls, using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). In addition, by using advanced statistical techniques, including machine learning methodologies, to analyze the data.The results indicated a significant link between both SNPs and DPN risk under both codominant and dominant models, respectively, with the A and T alleles as risk variants. Haplotype analysis further established the A-T haplotype as a prominent risk factor. The disease severity was associated with the 399 A/A and combined (G/A + A/A) genotypes, as well as the 194 C/T and combined (C/T + T/T) genotypes. In advanced DPN stages, random Forest (RF) highlighted both XRCC1 SNPs, and disease duration as the top three contributing factors. SHAP analysis corroborated the 194 C/T genotype of and the 399 A/A genotype were strongly linked to severe disease manifestations, particularly when coexisting with prolonged illness duration, advanced age, elevated HDL, and reduced LDL levels. Our findings substantiate the association of XRCC1 Arg399Gln and Arg194Trp SNPs with both susceptibility to and progression of DPN in T2DM patients. The integration of machine learning methodologies augments clinical decision-making by refining diagnostic precision and facilitating personalized treatment strategies.