308 Background: Reliable biomarkers are required to predict patient response to sorafenib. We attempted to investigate genomic variations associated with responsiveness to sorafenib treatment in patients with unresectable hepatocellular carcinoma (HCC) and their functional relevance. Methods: We obtained blood samples from 4 strong and 3 poor responders to sorafenib treatment and subjected these samples to whole-genome analysis. Next, we performed validation tests for candidate single-nucleotide polymorphisms (SNPs) in the samples of 174 HCC patients who were treated with sorafenib, followed by in vitro functional analysis and in silico analyses of candidate SNPs. Results: On average, 90 gigabases/sample was generated at ~34X sequencing depth. In total, 1813 genomic variations were perfectly matched to sorafenib responses in the clinical data; 708 were located within regions for sorafenib-target genes or drug absorption, distribution, metabolism, and excretion (ADME)-related genes—36 within the coding regions and 6 identified as non-synonymous single-nucleotide variants from 4 ADME-related genes (ABCB1, FMO3, MUSK, and SLC15A2), which potentially cause functional alterations. Validation tests of 174 patients confirmed sequencing results and revealed that patients with the C/C genotype for rs2257212 in SCL15A2 displayed higher risk for cancer progression than did patients with C/T or T/T genotypes (HR: 2.18; 95% CI, 1.15–4.15; P = 0.018). In vitro functional analysis revealed that cells harboring C/C genotype for this SNP displayed lower response to sorafenib treatment than did cells harboring the T/T genotype. Structural prediction analysis revealed change in protein phosphorylation levels, potentially affecting sorafenib-associated enzymatic activity. Conclusions: SLC15A2 could be a robust biomarker of response to sorafenib treatment in HCC patients.