Cancer of unknown primary (CUP) is a significant clinical problem accounting for 3-5% of all cancer diagnoses with an average patient survival of only 11 months(1,2). Contemporary CUP molecular tests can assist in identifying a likely tissue of origin, while the use of large gene panel clinical DNA sequencing can identify gene drivers of clinical or therapeutic importance. However, despite these recent advances our experience has been that many patients are currently not helped by either of these assays. Limitations include the amount and purity of tissue available (30-40% cases), low confidence or no match prediction of ToO (~30% cases) or failure to identify any pathogenic mutations (~20% cases). We are therefore now exploring newer methods to address these deficiencies. Firstly, we are using whole genome sequencing for problematic cases where gene-expression profiling and targeted DNA sequencing tests have not be useful. Secondly, we are developing methods to rapidly grow CUP organoid cell lines to expand tumour cells to circumvent issues with low tumour purity and acquiring sufficient nucleic acids for downstream molecular testing. Our longer term goal is to also use CUP organoids to predict drug response in patients. Finally, by exploring the immune landscape of CUP we observed that many CUP tumours have an inflamed tumour microenvironment with high expression of immunotherapy targets (e.g. high PD-L1 in ~30% of patients). Immune infiltration corresponds to specific tumour features including tumour mutation burden and viral aetiology meaning that these features may be useful for directing immunotherapy in CUP. Our ultimate goal is to use these combined approaches to provide a detailed molecular portraits of individual CUP tumours that will resolve their missing diagnosis and/or provide compelling evidence to guide precision treatment and improve patient outcome.