We read with great interest the recent article by Lu et al. [1] which assessed the association between polymorphisms of the ataxia telangiectasia mutated (ATM) gene and breast cancer risk with a meta-analysis involving five studies of 2,604 cases and 1,973 controls. This was an important investigation as the association has been inconsistent based on the previous individual studies. The results of metaanalysis indicated that for ATM 5557G[A polymorphism (also known as D1853N or rs1801516), AA genotype significantly decreased the risk of breast cancer compared with GA and GG genotypes [odds ratio (OR) = 0.67, 95% confidence interval (CI) 0.51-0.89]. However, another recent meta-analysis by Gao et al. [2] including nine studies of 4,191 cases and 3,780 controls suggested that ATM 5557G[A polymorphism was not significantly associated with breast cancer risk (AA vs. GA ? GG: OR = 0.78, 95% CI 0.59-1.04). After carefully reading the two articles, we noted several methodological issues in the study by Lu et al. [1].First, although the authors [1] claimed that they used comprehensive databases to search the potential eligible studies, they omitted at least five studies [3-7] that met their inclusion criteria for ATM 5557G[A polymorphism, which did make the strength of the association be inaccurate. However, as the study by Tapia et al. [3] was not in Hardy-Weinberg equilibrium (HWE) in controls, it should be excluded in the final meta-analysis. In addition, the meta-analysis by Gao L et al. [2] omitted one eligible study [8] and included one ineligible study [3]. Therefore, we re-assessed the association between ATM 5557G[A polymorphism and breast cancer (Fig. 1). Our results suggested that AA genotype was associated with decreased risk of breast cancer, compared with GA and GG genotypes in Caucasians (OR = 0.69, 95% CI 0.54-0.89).Second, there were several mistakes in Tables 1 and 2.