Chimeric antigen receptor T (CAR-T) cell therapy has produced outstanding results in clinical trials for hematological malignancies. However, this approach does not work for solid tumors as CAR-T cells may not be able penetrate tumor tissue. As an alternative to solve this limitation CAR natural killer (CAR-NK) cell therapy has emerged showing higher accessibility to the tumor, better safety, and 'off-the-shelf' manufacturing. Therefore, in this project we aim to develop a novel cellular therapy for the treatment of Head and Neck Squamous Cell Cancer (HNSCC) based on CAR-NK. Here we present our in silico and in vitro approach at the moment. We used a bioinformatic pipeline to identify targetable tumor-associated antigens using cancer and genomic databases and bioinformatic tools, finding twelve protein candidates classified into three tiers. After measuring the candidates in vitro expression levels in several HNSCC cell lines with real-time PCR (qPCR), two promising candidates where prioritize: a transmembrane adhesion receptor and a GPI-anchored cell membrane receptor. We designed several candidate single chain variable fragments (scFvs) directed to the target proteins and evaluated in silico predictions of the scFv-target interactions to select those with the highest affinity. Second-generation CAR-NK cells (4-1BB and CD3z costimulatory domains) were generated for both targets through lentiviral transfection of NK-92 cells with the plasmids containing the CAR constructs and assessed their cytotoxic activity in vitro. In the next step, we will compare our already generated CAR-NK cells with ones containing a NKG2D transmembrane domain and 2B4-CD3z costimulatory domains, typically present in NK cells.