The overall trend of cancer mortality in Japan has been decreasing since the 1960s (agestandardized death rates for ages 30-69), some 20-30 years earlier than in other industrialized countries. Cancer mortality was heavily influenced by Japanese postwar economic recovery, which led to improved living conditions and better control of some common forms of cancer (stomach, cervical) largely caused by infectious agents. However, Japanese wealth and development have also been associated with risky personal behaviors (smoking, drinking) and other conditions, leading to increases in cancers with no known or else very weak links to infection. We call this shift away from infectious and toward non-infectious causes of prevalent forms of cancers the "cancer transition," by analogy to Omran's "epidemiologic transition." The cancer transition described here in the case of Japan must be a part of efforts to revise and update the epidemiologic transition, which should incorporate new knowledge about the role of infection in chronic disease morbidity and mortality.