Reduced oxygen levels (hypoxia) is one of the most important factors influencing clinical outcome after radiotherapy. This is primarily because hypoxic cells are resistant to radiation treatment; hence, the greater the number of clonogenic cancer stem cells that exist under hypoxia, the lower the local tumour control. Reduced local control will influence overall survival, as may the hypoxic conditions by increasing malignant progression; however, to fight hypoxia, we should first be able to see it. We need noninvasive approaches that can accurately and reliably image hypoxia in tumours, especially using techniques that are routinely available in the clinic, such as PET, MRI and CT. All these imaging methods are already under clinical evaluation in this context. Such data should allow us to identify those patients on an individual basis who have hypoxic tumours and, thus, at the very least should receive some form of hypoxic modifier in conjunction with radiotherapy. Alternatively, the radiation dose could be either increased to the whole tumour or, if the imaging is accurate enough, only to the hypoxic subvolumes. The aim of this Review is to critically assess the potential use of imaging to help improve clinical outcome to radiotherapy.