This article reviews the current knowledge on the cancer-preventive potential of β-carotene, a precursor of vitamin A, and plentiful in fruits and vegetables, which has been studied widely as a promising chemopreventive agent in reducing the risk of cancer in humans. Several retrospective and prospective epidemiological investigations have demonstrated that a diet rich in micronutrients such as vitamins, carotenoids and selenium, could prevent the arising, in ‘high-risk’ patients, of precancerous and neoplastic lesions of specific sites, particularly of the upper aerodigestive tract. Numerous in vitro experiments have been performed in order to verify the true role played by this agent on cell proliferation and differentiation; until now, findings have been very encouraging, uniformly showing that β-carotene can affect carcinogenesis, particularly in early stages, through an antigenotoxic action. Antioxidant functions, immunomodulatory effects and control of intercellular messages via gap junctions are possible action mechanisms of the ability of β-carotene to block the carcinogenetic process. In vivo animal studies partially confirm the results obtained in vitro showing that β-carotene is able to reduce the induced cancer development; moreover, the association of the carotenoid with other microelements, such as vitamins E, C and glutathione often appears to be more effective than each agent used alone. From a clinical point of view, β-carotene appears an ‘ideal’ agent to be used in chemoprevention trials in humans, although optimal doses and intake methods need to be better defined; its almost zero toxicity permits the long-term administration of the drug, a vital condition for its anti-cancer activity, with good patient compliance. Human intervention studies performed so far, both randomized and uncontrolled clinical trials, have showed positive findings in specific cancer sites such as oral cavity, head and neck and colon; less consistent or negative are results on skin, lung and oesophagus cancer. The ongoing studies will provide more answers on these issues. A definitive evaluation of the ability of β-carotene to prevent cancer in humans requires further controlled trials; studies on a larger spectrum of cancer sites and different stages of disease must be encouraged. In addition, further investigation on biomarkers related to cancer risk and cancer incidence are necessary, particularly focused on the measurements for genotoxic damage, eg micronuclei, that may provide a valid and ‘easy’ marker for early stage carcinogenesis.