Genome-wide association studies have identified 40 ovarian cancer risk loci. However, the mechanisms underlying these associations remain elusive. In this study, we conducted a two-pronged approach to identify candidate causal SNPs and assess underlying biological mechanisms at chromosome 9p22.2, the first and most statistically significant associated locus for ovarian cancer susceptibility. Three transcriptional regulatory elements with allele-specific effects and a scaffold/matrix attachment region were characterized and, through physical DNA interactions, BNC2 was established as the most likely target gene. We determined the consensus binding sequence for BNC2 in vitro, verified its enrichment in BNC2 ChIP-seq regions, and validated a set of its downstream target genes. Fine-mapping by dense regional genotyping in over 15,000 ovarian cancer cases and 30,000 controls identified SNPs in the scaffold/matrix attachment region as among the most likely causal variants. This study reveals a comprehensive regulatory landscape at 9p22.2 and proposes a likely mechanism of susceptibility to ovarian cancer.Significance: Mapping the 9p22.2 ovarian cancer risk locus identifies BNC2 as an ovarian cancer risk gene. 12 16,915,387-16,915,739 NOTE: LD (r 2 ! 0.3) to rs3814113 based on 1000GP data v3. rs2153271 is reported in dbSNP as the reverse orientation to the genome. SNPs in bold represent the final five SNPs remaining at the end of the functional analysis. SNP in bold and underline indicates the only SNP that is common to the two analytic approaches. Abbreviations: MAF, minor allele frequency; NA, not available.Buckley et al.