Current evidence of chemopreventive activity for a variety of agents, from epidemiological studies and/or through laboratory investigations, is adequate to justify a range of clinical trials. These circumstances have prompted the International Agency for Research on Cancer to consider an International data evaluation programme in cancer prevention, analogous to the Monographs programme for carcinogens. Aspects of the Monographs programme that have contributed to its success merit examination in this context. It is clear that any new data evaluation programme must be grounded in a multidisciplinary assessment of data: relevant expertise extends from molecular genetics through to the design and evaluation of clinical trials. At the end, preventive efficacy for whole populations is unlikely: evaluations will probably indicate efficacy of an agent in relation to a particular organ or tumour type and in high-risk groups. Virtually all putative chemopreventive agents have adverse effects; many are carcinogenic. Such indications of hazard impinge upon the possible usage of these same agents to prevent cancer. Quantitative data concerning dose-specific effects, where available, will be relevant. It will also be important to elucidate the biological processes that account for chemoprevention and carcinogenesis - and to consider mechanisms relevant to other biological end points. Publication of evaluations of chemopreventive data may cause difficulty if such evaluations are misused or quoted out of context. This consideration and other complexities associated with the wider use of chemopreventive agents are discussed.