Chemoprevention can be defined as an attempt at cancer control in which the occurrence of the disease is prevented by the administration of one (or more) chemical compounds. Main problems in chemoprevention studies are the choice of a suitable drug, the choice of an appropriate intermediate or definitive end point, and the definition of the population which should be investigated. Main classes of chemopreventive agents include vitamins, non-steroid antinflammatory drugs, minerals such as calcium or selenium, and other antioxidants such as N-acetylcysteine. Chemoprevention is particularly appealing in colorectal cancer, either because these lesions develop through a multistep process, or owing to the concept of "field carcinogenesis'. Between 1985 and 1990 we carried out a controlled study in which antioxidant vitamins or lactulose were used in an attempt to prevent the recurrence of colorectal polyps after their endoscopic removal. Among the 209 patients who could be evaluated, polyps recurred in 5.7% of the individuals who were given vitamins (A, C and E), 14.7% of patients given lactulose and 35.9% of untreated controls (chi 2 = 17.1, P < 0.001). The study suggested that either antioxidant vitamins or lactulose could be effective in reducing the recurrence rate of adenomatous polyps. In a subsequent on-going study, lower doses of the same vitamins were tested versus N-acetylcysteine (600 mg/day) or no treatment. Preliminary results showed a 40% reduction of the recurrence of polyps (versus controls) in individuals given N-acetylcysteine, while the effect of lower doses of vitamins was less appreciable. Definitive results of the study should be available by the end of 1998.