Oxidative DNA damage is widely believed to play a role in cancer aetiology. It is therefore important to be able to assess it, both as an index of cancer risk, and in experiments to test agents with a potential to reduce oxidative damage, such as dietary antioxidants. However, there is an alarming discordance in estimates of concentrations of oxidative damage in human DNA, largely attributable to the kind of method used to measure it. A meeting was held recently at the Rowett Research Institute in Aberdeen to address this problem.