Epidemiological studies showing a protective effect of diets rich in fruits and vegetables against cancer have focused attention on the possibility that biologically-active plant secondary metabolites exert anti-carcinogenic activity. This huge group of compounds, now collectively termed 'phytochemicals', provides much of the flavour and colour of edible plants and the beverages derived from them. Many of these compounds also exert anti-carcinogenic effects in animal models of cancer, and much progress has been made in defining their many biological activities at the molecular level. Such mechanisms include the detoxification and enhanced excretion of carcinogens, the suppression of inflammatory processes such as cyclooxygenase-2 expression, inhibition of mitosis and the induction of apoptosis at various stages in the progression and promotion of cancer. However, much of the research on phytochemicals has been conducted in vitro, with little regard to the bioavailability and metabolism of the compounds studied. Many phytochemicals present in plant foods are poorly absorbed by human subjects, and this fraction usually undergoes metabolism and rapid excretion. Some compounds that do exert anti-carcinogenic effects at realistic doses may contribute to the putative benefits of plant foods such as berries, brassica vegetables and tea, but further research with human subjects is required to fully confirm and quantify such benefits. Chemoprevention using pharmacological doses of isolated compounds, or the development of 'customised' vegetables, may prove valuable but such strategies require a full risk-benefit analysis based on a thorough understanding of the long-term biological effects of what are often surprisingly active compounds.