Background: The Cancer Genome Atlas and others identified genomic events suggesting that endometrial cancer (EC) should be susceptible to DNA repair inhibition. Mutations in classical homologous recombination genes occur in 22% of EC, ARID1A 41% and PTEN loss occurs in 55% of EC. Data from pre-clinical models suggest poly ADP-ribose polymerase (PARP) inhibitors alone or in combination may be an effective therapeutic strategy in EC (Hansen 2016). Combinations of angiogenic inhibitors and PARP inhibitors have demonstrated synergistic effects and have been well tolerated in other tumor types. This study has been designed to compare 2 experimental arms exploring DNA repair inhibition versus cediranib alone which has previously shown promising activity in GOG 229J (Bender 2015). Methods: This is a multicenter randomized three arm study for patients with recurrent, metastatic or persistent EC. Patients are randomized 1:1:1 to cediranib PO 30 mg OD; olaparib 300 mg PO BID or the combination of cediranib 20 mg PO OD with olaparib 300 mg PO BID. All treatment cycles are 28 days. Primary endpoint is progression free survival (PFS). The study is powered to detect an increase in median PFS from 3.6 (based on cediranib alone) to 7.2 months with 90% power, using a one-sided test with α = 0.05 per comparison. Forty patients will be enrolled per arm, with an interim futility analysis planned. Eligibility includes endometroid, serous, and mixed histology EC; at least 1 prior line of chemotherapy (no more than 2 lines for metastatic disease), prior endocrine or immunotherapy is allowed; ECOG PS ≤ 2; adequate hepatic, bone marrow, coagulation and renal function. Archival tumor tissue and blood samples are being collected for translational studies. The study is open across the NRG network; 24 patients are enrolled to date . Amendments are planned to include additional arms investigating combination strategies targeting DNA repair and angiogenesis. Clinical trial information: 03660826.