High-grade epithelial ovarian carcinomas (OC) containing mutated BRCA1/2 have homologous recombination defects and are sensitive to poly(ADP-ribose) polymerase inhibitors (PARPi). In a clinical trial of the PARPi rucaparib (ARIEL2 Part 1, Clovis Oncology) a patient was observed with a germline truncating mutation in RAD51D (c.770_776del, p.G258Sfs*50) and a secondary mutation (c.770_776delinsA, p.S257_R259delinsK) in a biopsy of a splenic lesion that was progressing on PARPi therapy. Evolutionary analysis and molecular dynamics modelling were used to assess the function of this variant of unknown effect alongside the functional wild-type variant(s). Results indicated that the observed differences in amino acid sequence between the secondary mutation and wild-type RAD51D were unlikely to disrupt normal function and are evolutionarily well tolerated. The secondary mutation (c.770_776delinsA, p.S257_R259delinsK) would likely mirror the function of wild-type RAD51D, thus would restore function and lead to PARPi resistance. This prediction was confirmed by CRISPR directed homology repair introduction of the secondary mutation into a human ovarian cancer cell line, PEO4, which demonstrated a decreased cisplatin and rucaparib sensitivity relative to a PEO4 RAD51D knockout. In conclusion, the secondary RAD51D mutation (c.770_776delinsA, p.S257_R259delinsK) identified in this lesion most likely contributed to or caused the PARPi resistance and lesion progression.