Objective: The study aimed to ascertain the correlation between AKR1B1 polymorphism rs759853 and the risk of diabetic retinopathy (DR) through a meta-analysis.;-;Methods: Crude odds ratios (ORs) and the corresponding 95% confidence interval (95% CIs) were calculated to assess the association of AKR1B1 rs759853 polymorphism with DR risk. Stratification analyses were further conducted based on ethnicity, diabetes mellitus (DM) type, Hardy-Weinberg equilibrium (HWE) status, and genotyping method. Heterogeneity was detected by Q test. Sensitivity analysis was implemented to check the robustness of final results. Additionally, Begg's funnel plot and Egger's test were used to evaluate underlying publication bias.;-;Results: Our meta-analysis ultimately incorporated 21 eligible publications with 22 independent case-control studies. The overall results demonstrated that AKR1B1 rs759853 polymorphism had no association with DR risk under all genetic models. However, after subgroup analysis by DM type, the rs759853 polymorphism was a protective factor against the DR onset in patients with type 1 DM (TT vs. CC: OR = 0.33, 95% CI = 0.17-0.67; TT + CT vs. CC: OR = 0.49, 95% CI = 0.36-0.68; TT vs. CC + CT: OR = 0.48, 95% CI = 0.28-0.83; allele T vs. allele C: OR = 0.56, 95% CI = 0.44-0.72; CT vs. CC: OR = 0.52, 95% CI = 0.37-0.74). Furthermore, subgroup analysis by genotyping method suggested that rs759853 genotyped using MassARRAY assay was significantly correlated with decreased risk of DR under dominate model (TT + CT vs. CC: OR = 0.71, 95%CI = 0.52-0.96).;-;Conclusion: AKR1B1 polymorphism rs759853 may inhibit the occurrence of DR in patients with type 1 DM.