Epidemiological studies on the association between the single nucleotide polymorphism (SNP) at -420 C/G (rs1862513) in the human resistin gene (RETN) and the risk of type 2 diabetes mellitus (T2DM) are conflicting. In order to derive a more precise estimation of the association, a meta-analysis was conducted. Twelve studies with 5,935 cases and 5,959 controls were enrolled by searching the databases of Medline, EMBASE, and Cochrane. Summary odds ratios (ORs) with 95% confidence intervals (CIs) were calculated. The heterogeneity and publication bias were investigated. The main analysis indicated no significant association [for allelic model: OR = 0.97 (0.92-1.03); for additive model: OR = 0.95 (0.83-1.09); for recessive model: OR = 0.98 (0.86-1.12); for dominant model: OR = 0.95 (0.88-1.04)]. Overall, no significant heterogeneity was found. Subgroup analysis by race and source of controls indicated no significant association. In conclusion, the current meta-analysis did not observe any association between the polymorphism of RETN -420 C/G and the risk of T2DM. The study may help us further understand the genetics of T2DM. However, larger and prospective studies are warranted to confirm this finding.