One of the challenges of studying anticancer therapies is that effects observed in cell lines or mouse models are not always good indicators of clinical trial results. The mouse model of acute promyelocytic leukaemia has bucked this trend, as targeted therapies such as retinoic acid and arsenic induce differentiation and clearance of leukaemia cells in both mice and humans. This mouse model has also provided important mechanistic insights into the combinatorial effects of these agents and has promoted combined therapies that have shown recent success in the clinic.