Dietary guide-lines are currently given to whole populations whether the recipients have asked for it or not, and this raises ethical and practical problems. If we have not been asked for advice, what right do we have to tell people what they "ought" to eat? Are we sufficiently sure of the benefits of following our advice? Could we, in fact, be wrong about it? Are we sure of the absence of "side effects" of our advice? If not, are we sure that the balance between risk and benefit is favourable? Might our advice, in fact, do more harm than good? Would it not be better to restrict our advice to those at high risk of diet- related cancer? These issues are all discussed. Furthermore, the conclusions from the European Cancer Prevention (ECP) symposium and workshop on public education on diet and cancer are presented. The major conclusion is that the concept of general guide-lines should be abandoned in favour of guide-lines targeted to specific groups. In addition, it is concluded that the recent evidence had led to considerable confusion regarding dietary fibre and cancer, and also regarding dietary fat and cancer. The recent evidence strengthens the association between high intake of fresh fruit and vegetables and low risk of cancers at a number of sites.