Colorectal cancer (CRC) remains a cause of significant mortality in developed countries despite extensive knowledge of its epidemiology and molecular basis. Since multiple molecular steps that collectively bring about this disease are known, its chemoprevention is a realistic proposition. Biochemical targets of CRC chemopreventive agents include carcinogen metabolising enzymes, arachidonic acid metabolism, the transcription factor nuclear factor-kappa beta (NF-kappaB), enzymes responsible for polyamine metabolism, and events associated with proliferation and apoptosis of preneoplastic cells. Aspirin, celecoxib, calcium and alpha-difluoromethylornithine are examples of drugs that have undergone clinical testing. Critical evaluation of these trials allows optimisation of methodologies for clinical advancement of novel chemopreventive agents. Cancer patients can be a suitable cohort of subjects for pilot studies of certain new agents. Such studies and larger trials in high-risk healthy individuals require the stringent use of carefully validated 'preneoplastic' biomarkers which are intrinsically related to defined stages of colorectal carcinogenesis and/or to mechanisms of action of the agent under investigation.