Although several hypotheses for human carcinogenesis have been proposed, the specific genetic changes that cause normal cells to become cancer cells have not been identified. In spite of uncertainties regarding the mechanisms of carcinogenesis, several vitamins such as beta-carotene and vitamins A, C, and E, which can reduce the risk of cancer, have been identified, using animal and in vitro models of carcinogenesis. These studies have led to a hypothesis that the supplemental intake of these vitamins may reduce the risk of cancer. This hypothesis in humans can be tested only by intervention trials that are in progress. Prospective and retrospective case-controlled experimental designs are not suitable for testing the above hypothesis. The fact that some vitamins induce cell differentiation and/or growth inhibition in tumor cells in culture suggests that the use of these vitamins in cancer prevention has a cellular basis. In addition to having a direct effect on tumor cells, vitamins such as alpha-tocopheryl succinate and beta-carotene enhance the effect of other agents that induce differentiation in tumor cells. Some vitamins like beta-carotene, retinoic acid, alpha-tocopheryl succinate, and vitamin D also regulate the expressions of certain oncogenes and cellular genes. These are exciting new functions of vitamins that nobody could have predicted only a few years ago.