BACKGROUND : Alternative splicing regulates most of protein-coding genes by producing diverse messenger RNA transcripts; and mis-splicing events can induce aberrant protein isoforms that contribute to cancer development. It is possible that genetic variations in splicing associated genes may regulate the formation of transcripts and multiple protein isoforms by affecting the splice regulatory elements. In this study, we aimed to determine whether genetic variations in the crucial alternative-splicing genes were associated with breast cancer risk. METHODS : A case-control study was conducted with 1064 breast cancer cases and 1073 healthy controls from China. A total of 16 tagging polymorphisms within three splicing factor-associated genes (SFRS3, ESRP1 and ESRP2) were genotyped by using Infinium BeadChip. The association between the polymorphisms and risk of breast cancer was evaluated by computing odds ratios (OR) and 95% confidence intervals (CIs). RESULTS : The genotype distribution of rs2145048 in SFRS3 was different between cases and controls (Bonferroni corrected P = 0.022). After adjusting for age, age at menarche and menopausal status, the A allele of rs2145048 showed an inverse association with breast cancer risk in the additive model (adjusted OR = 0.81, 95% CI = 0.71-0.92, P = 0.001, Bonferroni corrected P = 0.016). In the stratification analysis, the association between rs2145048 A allele and breast cancer remained significant in subgroups of earlier menarche, older first born, premenopausal status, and ER/PR negative status. CONCLUSIONS : This study provided the first evidence that SFRS3 rs2145048 was associated with breast cancer susceptibility in Chinese women, which might represent a biomarker to improve the identification of individuals at high risk of this malignancy.