Widening indications for combining radiation therapy with cytotoxic or molecular-targeted drugs have mainly been driven by pragmatic clinical trials. With a flurry of novel drugs in various stages of preclinical and clinical development there is a need to revise the framework that has traditionally been used for discussing possible drug-radiation interactions, especially because many of the new drugs are directed at a specific molecular target. Spatial cooperation, cytotoxic enhancement, biological cooperation, temporal modulation and normal tissue protection are proposed as five primary exploitable mechanisms for the rational combination of drugs with radiation for cancer therapy. These five mechanisms produce different clinical outcomes and, therefore, the optimum clinical end point for assessing therapeutic benefit will depend on the mechanism tested. The dependence of outcome on these mechanisms also affects the selection of preclinical models and the optimum scheduling of the two modalities, i.e. the timing and dosing of the drug relative to the radiation dose fractions. These considerations are discussed in some detail for each mechanism and illustrated with specific clinical examples. Multi-modality therapy for head and neck squamous-cell carcinoma is used to illustrate these concepts. Further clinical progress in this field will require hypothesis-driven trials to ensure efficient identification of treatments with the most favorable risk:benefit ratio.