The role of mitochondrial DNA (mtDNA) variant 16189T>C in type 2 diabetes mellitus (T2DM) remains hotly debated in the past decade. If mutation 16189T>C indeed posed a risk to T2DM, as echoed by some recent studies, correlation between this mutation and disease should be observed when carrying out a systematical study using data and samples collected in a large geographic region in China. To test this hypothesis, we first performed a linear regression analysis between the prevalence of T2DM and the allele frequency of 16189C variant in 10 East Asian populations, and further genotyped this variant in two casecontrol cohorts from west Han Chinese (Kunming and Xining). Linear regression analysis showed that no significant correlation was observed (r2=0.211, P=0.181), and the genotyping results indicated that the m.16189T>C frequency difference between case and control was not significant in either populations (P=0.38 and 0.89 for Kunming and Xining, respectively). Matrilineal backgrounds constitution (in terms of haplogroups) analysis generated a similar haplogroup distribution in both populations (P>0.1). All results failed to substantiate that m.16189T>C may play an active role in the development of T2DM in East Asian populations.