Background: Homologous recombination repair (HRR)-deficient Ovarian Cancer (OvC) is sensitive to PARP inhibitors and platinum salts. Genomic HRR-deficiency (HRD) tests (i.e. Myriad MyChoice) are currently used to guide treatment selection in the maintenance setting. However, they are showing a limited predictive value, they are expensive and static, i.e. unable to capture the restoration of the HRR functionality, as a mechanism of drug resistance. Methods: We performed the functional and dynamic RAD51 test on 164 patients with high-grade serious OvC, enrolled in the MITO16A/MANGO OV2 trial and treated with first-line therapy containing carboplatin, paclitaxel, and bevacizumab. The RAD51 assay included the immunofluorescence of gH2AX (as a marker of Double Strand Breaks DNA damage), BRCA1 and RAD51 nuclear foci. Stromal Tumor Infiltrating Lymphocytes (TIL) by Immunohistochemistry, and gene expression profile were performed on the same samples. Results: RAD51 assay was informative on 70% (115/164) of the samples. On evaluable samples, median gH2AX score was 57%. BRCA1 foci by immunofluorescence were detected on 55% (49/112) of the samples; 45% (63/112) of the tumors did not show BRCA1 foci likely due to germline or somatic BRCA1 mutations, or BRCA1 promoter hypermethylation. RAD51 identified 69% (79/115) of all tumors as HRD, with a median RAD51 score of 8.8%. HRD tumors presented a statistically significantly higher TIL content than HRR-proficient ones (p = 0.03). Conclusions: Our preliminary results confirmed the feasibility of the RAD51 functional assay on primary untreated OvC. The association between HRD by RAD51 and TIL content encouraged to investigate the efficacy, in clinical trials, of the combination of PARPi and immune-checkpoint inhibitors on selected patients (i.e. HRD by RAD51 and TIL high). Gene expression profile and comparison of the HRD status by RAD51 and genomic HRD (both Myriad MyChoice and academic) and correlations with clinical outcomes will be available for the meeting.