Epithelial ovarian cancer (EOC) has a substantial hereditary component. Germline mutations in the high- penetrance susceptibility genes BRCA1 and BRCA2 confer EOC risks of around 50% by age 70. These genes are responsible for most families containing several cases of EOC; but the known susceptibility genes account for less than 40% of the excess familial risk of EOC indicating that other genes await discovery. We hypothesize that a substantial proportion of the excess EOC risk is due to moderately penetrant, rare or uncommon functional variants in coding DNA conferring 2 to 10 fold lifetime risks of disease. Identifying additional EOC susceptibility genes could have a major and rapid clinical impact by reducing disease associated mortality through better risk prediction-prevention strategies, and developing novel therapeutic approaches and individualized treatments. There are three phases to this project: The first is a discovery phase, in which we plan to perform exome sequencing in ~200 non-BRCA1/BRCA2 EOC families to characterize the genome wide spectrum of functional coding mutations in these families. We will also leverage sequencing data from ~400 EOC cases that have undergone germline exome sequencing through The Cancer Genome Atlas (TCGA) project, and another ~150 EOC families for whom exome sequencing is ongoing. We will then compare the frequency of candidate functional variants we find with data from a combined exome sequence analysis of ~16,000 non-cancer subjects of European origin from multiple different studies. Next, is a replication phase in which we plan to follow-up candidate genes and variants identified in phase 1, in large-scale sequencing and genotyping efforts, leveraging the epidemiological case-control collections of an international collaboration, the Ovarian Cancer Association Consortium (OCAC). We will perform candidate gene next generation sequencing for approximately 100 genes in 5,000 EOC cases and 5,000 controls to characterize the prevalence of functional deleterious mutations; and for ~20,000 non-synonymous rare variants we identify we will use a customized array to genotype these variants and an additional 260,000 rare variants from the 11,000 exome sequencing project, in 5,000 EOC cases and 5,000 controls to evaluate their disease associations. In the final phase, we will perform penetrance and survival analyses for confirmed functional variants to evaluate their clinical significance, which will indicate the immediate clinical value of the genes/variants we identify.