Background: Tamoxifen is a selective estrogen receptor modulator that is widely used to treat estrogen receptor (ER)-positive breast cancer. However, not all patients benefit from the incorporation of tamoxifen into an adjuvant therapy. This is also the case when tamoxifen is used in chemoprevention, since only half of participants benefit from the drug. In order to improve treatment response, we attempted to identify single nucleotide polymorphisms (SNPs) that correlated with tamoxifen efficacy. Methods: ER-positive breast cancer patients at our hospital were enrolled on this study between January 2007 and September 2010. The primary endpoint was ER-positive breast cancer-free survival. We examined 17 SNPs in these patients. The survival benefit associated with each genotype was determined with a log-rank test, and the hazard ratio was analyzed using a Cox proportional-hazards model. Results: The median follow-up time of the 320 patients enrolled on the study was 3298 days. Of 240 patients who received any endocrine therapy, ER-positive breast cancer-free survival in patients with the 2q35 rs13387042 AA genotype was significantly shorter than in those who had the AG or GG genotype (p < 0.0001), and the hazard ratio was significantly higher (HR 8.83; 95% CI 2.09–25.53, p = 0.0064). Of the 145 patients who received tamoxifen therapy, there was a trend among ER-positive breast cancer patients with the CYP2C19 rs4917623 TT genotype to have a shorter disease-free period (p = 0.0635) when compared to patients with TC or CC genotypes. Similarly, there was a trend for the TT genotype patients to exhibit a higher hazard ratio (HR 2.62; 95% CI 0.86–7.55, p = 0.0861). Conclusions: The rs4917623 SNP in the CYP2C19 gene, which encodes a metabolic enzyme, predicts tamoxifen efficacy. This finding will facilitate selection of ER-positive breast cancer patients for tamoxifen treatment; it may also be useful for selection of patients most likely to benefit from tamoxifen-dependent chemoprevention.