Interventions designed to delay or prevent the development of invasive colorectal adenocarcinoma might result in a profound impact on the incidence of and mortality from colorectal cancer. Current developmental efforts focus in two major directions: identification and testing of combination-targeted signal transduction pathway modulators and development of nutriceuticals. Both approaches recognize the heterogeneity of the carcinogenesis process. Combinations of agents (eg, aspirin or sulindac with DFMO, calcium with fiber), which have complementary or synergistic mechanisms or agents with broad spectrum anti-carcinogenic effects can potentially improve upon the effects of single agents and account for redundant signal transduction pathways involved in uncontrolled growth of clonal cells. Preliminary data in rodent systems suggest nutriceuticals or standardized foodstuffs, which contain multiple components, many of which are not identified have synergistic anticarcinogenesis effects. Diet modulation or supplementation with known standardized foodstuffs may be ideal in individuals at risk for colorectal cancer. Current clinical models for colorectal cancer prevention focus upon a pathologic-surrogate endpoint (reducing the recurrence or emergence of adenomas). This surrogate, generally accepted at a regulatory level, might not represent the true malignant progression from normal colonic mucosa to the malignant phenotype. Could the intervention simply prevent the emergence of adenomas that would not have progressed to invasive neoplasia in any event? Despite this major weakness, the adenoma remains the regulatory endpoint for efficacy because the cancer event remains rare, even among individuals selected as high-risk epidemiology but without known, highly pressured genetic-based stress. Over the next decade, the emergence of molecular and proteomic profiling tools have the potential of selecting adenomas that will progress to adenocarcinomas and will allow for investigation to be focused on individuals at increased risk.