The term 'translational research' is currently in vogue, but what does it mean? Faced with patients in the clinic everyday, translational research could mean the use of science to improve the quality of life of patients with cancer, or to prevent the over treatment of others. However, the view from the laboratory bench might be somewhat different — translational research at best might be considering the relevance of basic research to current clinical problems, at worst, just another essential grant funding phrase. But, to ensure improved treatment for patients with cancer, do we need to look more carefully at translational research?

These are some of the questions debated in 'What is translational cancer research?', our current cancer podcast (http://www.nature.com/nrc/podcast/translational.html). Our panel, made up of both clinical and basic research scientists, are clear on one thing; that a greater degree of integration between scientists and clinicians is mandatory if we are to improve the translation of science into the clinic, and that this is definitely a two-way street.

We also need more effective tools for translation between the laboratory and the clinic. One way in which we might achieve this is through a critical review of how we use and interpret data from genetically engineered mouse models of cancer, and how we relate these to results from the clinic and from the culture flask. The benefits and pitfalls of using mouse models to better understand the complexity of human tumours is the subject of three of this month's articles (see pages 645, 659 and 707), and of our poster on 'The evolution of genetically engineered mouse models of cancer', which is freely available online at http://www.nature.com/nrc/posters/mousemodels.

We are grateful to the Cancer Research UK Cambridge Research Institute for sponsoring both our podcast and our poster.