Cancer “stem cells” (CSCs) sustain the hierarchies of dividing cells that characterize cancer. The main causes of cancer-related mortality are metastatic disease and relapse, both of which originate primarily from CSCs, so their eradication may provide a bona fide curative strategy, though there maybe also the need to kill the bulk cancer cells. While classic anti-cancer chemotherapy is effective against the dividing progeny of CSCs, non-dividing or quiescent CSCs are often spared. Improved anti-cancer therapies therefore require approaches that target non-dividing CSCs, which must be underpinned by a better understanding of factors that permit these cells to maintain a stem cell-like state. During hematopoiesis, retinoic acid receptor (RAR) γ is selectively expressed by stem cells and their immediate progeny. It is overexpressed in, and is an oncogene for, many cancers including colorectal, renal and hepatocellular carcinoma, cholangiocarcinomas and some cases of acute myeloid leukemia that harbor RARγ fusion proteins. In vitro studies suggest that RARγ-selective and pan-RAR antagonists provoke the death of CSCs by necroptosis and point to antagonism of RARγ as a potential strategy to treat metastatic disease and relapse, and perhaps provide a cure for some cancers.