Hemophagocytic lymphohistiocytosis (HLH) is a hyperinflammatory disorder characterized by a life-threatening cytokine storm and immunopathology. Familial HLH type 3 (FHL3) accounts for 30% of all inborn HLH cases worldwide. It is caused by mutations in the UNC13D locus, which result in impaired processing of cytotoxic vesicles and hence compromised T- and NK cell-mediated killing. Current treatment protocols, including allogeneic hematopoietic stem cell (HSC) transplantation, still show 30-40% mortality. This unmet medical need warrants the exploration of curative genome editing strategies. The Jinx mouse model, which harbors a cryptic splice site (cSD) in Unc13d intron 26, reflects human FHL3 very well. Here, we employed CRISPR-Cas technology to excise the cSD in HSCs. For validation, Jinx HSCs were edited with CRISPR-Cas9 ribonucleoprotein complexes (RNPs) and transplanted into conditioned Jinx recipients. Flow cytometry confirmed >95% engraftment and successful establishment of a donor-derived immune system. Genotyping based on NGS, CAST-Seq and ddPCR performed on genomic DNA of cells isolated from transplanted mice confirmed efficient gene editing (>95%) and absence of off-target effects. Unc13d transcription levels of edited and non-edited cells were comparable. Challenge of the transplanted mice with lymphocytic choriomeningitis virus (LCMV) resulted in rapid virus clearance and protection from HLH in Jinx mice transplanted with edited HSCs while Jinx mice transplanted with mock-edited cells developed a full blown HLH. In sum, we established a highly efficient CRISPR-Cas RNP-based editing protocol for murine HSCs, which allowed us to fully restore the functionality to cytotoxic T cells in a preclinical familial HLH disease model.