Ultraconserved elements (UCEs) are the most extreme representatives of conserved non-coding sequences. Recent studies have indicated that UCEs are not mutation cold regions and likely to be concerned with cancers, including breast cancer (BC). In this study, we first screened common single-nucleotide polymorphisms (SNPs) (minor allele frequency, MAF > 0.05) in Chinese population located in 481 UCEs sequences and selected seven SNPs (rs17049105, rs13020355, rs2682406, rs2056116, rs11190870, rs9572903, and rs8004379) of uc.51, uc.82, uc.133, uc.140, uc.302, uc.353, and uc.368, respectively. A two-stage case-control study of BC with a total of 1,497 cases and 1,497 controls in Chinese population was conducted to test the hypothesis that these SNPs of UCEs are associated with BC risk. Stage I with 735 cases and 735 controls was designed to discover the risk variants, followed by stage II with 762 cases and 762 controls to validate the significant variants. In stage I, although the genotype distributions of all seven SNPs were not significantly different between BC cases and controls, logistic regression analyses revealed that the variant genotypes of rs8004379 were significantly associated with the increased risk of BC (dominant model: adjusted OR = 1.27, 95% CI = 1.01-1.58, P = 0.039). We then selected two SNPs, rs8004379 A/C and rs2056116 A/G, with lowest P values of the associations into the stage II analysis. However, none of above two SNPs were significantly associated with BC risk in both stage II and pooled set (rs8004379 AC/CC vs. AA: adjusted OR = 0.88, 95% CI = 0.68-1.13 for stage II and adjusted OR = 1.09, 95% CI = 0.92-1.29 for the pooled set; rs2056116 AG/GG vs. AA: adjusted OR = 1.12, 95% CI = 0.87-1.45 for stage II and adjusted OR = 1.11, 95% CI = 0.94-1.31 for the pooled set). These findings did not support a significant association between UCEs SNPs and the risk of BC in Chinese population.