<b>Objectives:</b> To evaluate differences in rates of pathogenic/likely pathogenic germline variants (P/LPV) and subsequent genetic counseling by self-reported ancestry in epithelial ovarian cancer (EOC) patients (pts) undergoing tumor-normal sequencing for cancer care. <b>Methods:</b> EOC pts were prospectively consented to tumor-normal sequencing via a custom NGS panel (MSK-IMPACT) from January 2015 to December 2019, inclusive of germline analysis up to 88 genes. Designated EOC-associated genes included: <i>BRCA1 BRCA2, RAD51C, RAD51D, BRIP1, PALB2, ATM, MLH', MSH2, MSH6,</i> and <i>PMS2.</i> Variants of uncertain significance were not included. All women with new P/ LPV were referred to a Clinical Genetics Service (CGS) for counseling. Ancestry was defined using self-reported definitions of race/eth- nicity and designations of Ashkenazi Jewish (AJ) ancestry. Pts were categorized into mutually exclusive groups (AJ, Non-Hispanic [NH]- White, Non-White [Asian, Black/African American, Hispanic, other], or unknown). Patients identifying as AJ or Hispanic were classified as such, regardless of race. Clinical characteristics were analyzed using summary statistics. Associations between ancestry, mutation status, and CGS follow-up were assessed with non-parametric statistical tests (STATA V17). <b>Results:</b> Of 919 pts with EOC, 235 (25.6%) had a P/LPV, with 149 (16.2%) in an EOC-associated gene. Of those with P/LPV, 91 (38.7%) previously knew about this result for all genes, and 84 (56.4%) knew of this result for EOC-associated genes. For those with new findings, 80% presented for CGS counseling. The median age of EOC diagnosis for the entire cohort was 60 years (range: 16-90). Carriers of a P/LPV had a younger age at diagnosis compared with pts with sporadic EOC (59 vs 61, p=0.012). High-grade serous EOC (HGSOC) comprised 77% of cancers, and P/LPV carriers were more likely to have HGSOC than other histologies (85.5% vs 74.1%, p<0.001). Women were identified as NH-White (58%), AJ (17%), Asian (10%), Hispanic (6%), Black (5%) and unknown (4%) ancestries. The prevalence of P/LPV differed by ancestry for all genes (NH-White 22%, AJ 41%, Asian 28%, Hispanic 18%, Black 21%, unknown 21%, p<0.001). When assessing specifically for EOC-associated genes, a similar pattern was observed; AJ pts had the highest prevalence of P/LPV (25.6%). After removal of AJ pts, rates of EOC-associated P/LPV differed between NH-White (12%), Asian (25%), Hispanic (18%), and Black (21%) pts (p=0.021). There were no significant differences between ancestry groups in rates of prior knowledge of P/LPV or CGS follow-up rates for newly diagnosed pts with P/LPV. <b>Conclusions:</b> Rates of P/LPV, particularly in EOC-associated genes, were high in Non-White pts, particularly in Asian pts, compared with Non-AJ, NH-White pts. Rates of follow-up counseling were similar between ancestry groups. These findings highlight the need for universal genetic testing in all EOC pts regardless of ancestry, given the implications for treatment and cascade testing.