The epidemiologic findings for intestinal cancer are reviewed and a model that may fit most of the facts is proposed for further testing. The major features of the model for colon cancer are: a. In low-risk populations where the disease is "endemic," colon cancers are concentrated in the cecum and ascending colon, female cases are preponderant, and most of the rise to the maximum incidence level has occurred by age 50–55; b. When a new etiologic factor is introduced into such a population, the transition from an "endemic" to an "epidemic" phase is first expressed as a rise in sigmoid cancers among older men over 55 years; c. A rise in female sigmoid cancers follows later, and the time lag is reinforced by a tendency for these female cases to appear at somewhat older ages than the male cases. Some tests of the model and implications for studies of suspect precursor lesions such as adenomatous polyps are indicated. In the latter connection, the need for comparative studies among populations subject to widely differing colon cancer risks is emphasized.