Despite the identification of several ovarian cancer (OC) predisposition genes, a large proportion of familial OC risk remains unexplained. We adopted a two‐stage design to identify new OC predisposition genes. We first carried out a large germline whole‐exome sequencing study on 158 patients from 140 families with significant OC history, but without evidence of genetic predisposition due to BRCA1/2. We then evaluated the potential candidate genes in a large case–control association study involving 381 OC cases in the Cancer Genome Atlas project and 27,173 population controls from the Exome Aggregation Consortium. Two new putative OC risk genes were identified, namely, ANKRD11, a putative tumor suppressor, and POLE, an enzyme involved in DNA repair and replication. These two genes likely confer moderate OC risk. We performed in vitro experiments and showed an ANKRD11 mutation identified in our patients markedly lowered the protein expression by compromising protein stability. Upon future validation and functional characterization, these genes may shed light on cancer etiology along with improving ascertainment power and preventive care of individuals at high risk of OC.