ObjectiveTo systematically evaluate the association between eNOS gene a/b polymorphism and diabetic retinopathy susceptibility. MethodsPubMed, EMbase, The Cochrane Library (Issue 5, 2015), CBM, CNKI, VIP and WanFang Data were systemically searched from inception to May 2015, to collect case-control studies about the eNOS a/b polymorphism and diabetic retinopathy susceptibility. Two reviewers independently screened literature, extracted data and evaluated the risk bias of included studies. Then, meta-analysis was performed by RevMan 5.2 software. ResultsA total of 16 case-control studies were included, which involved 3 232 diabetic retinopathy cases and 3 555 controls. The results of meta-analysis showed that, in the total analysis, the a/b polymorphism of the eNOS gene was not associated with diabetic retinopathy risk (dominant model:OR=0.94, 95%CI 0.78 to 1.15, P=0.57; recessive model:OR=0.97, 95%CI 0.78 to 1.22, P=0.88; aa vs. bb:OR=0.89, 95%CI 0.71 to 1.12, P=0.32; ab vs. bb:OR=0. 94, 95%CI 0.77 to 1.14, P=0.52; a vs. b:OR=0.97, 95%CI 0.82 to 1.14, P=0.70); In the subgroup analysis by ethnicity, the a/b polymorphism was significantly associated with the risk of diabetic retinopathy in Africans, but not in Asians and Caucasians. ConclusionThe a/b polymorphism in the eNOS gene may be a risk factor of diabetic retinopathy in Africans.