162 The length of human life, usually measured collectively by the life expectancy at birth (e0), increased substantially in high-income countries during the twentieth century, and has so far been rising further in this century. The pace of progress was particularly fast in Japan, where e0 reached 86.44 for females and 79.59 for males in 2009. The increasing life expectancy in Japan during the last few decades led to an accelerating growth of the extremely old population (Robine, Saito and Jagger, 2003). The progress in Japan has been faster than those in most Western high-income countries (Figure 1). Why has Japan achieved such a rapid longevity extension? Previous studies pointed out such factors as progress in medical technology and development of the national medical insurance system (Ohno, 1985). These factors, however, are observed in many other high-income countries as well. Thus this paper focuses on factors that may be peculiar to, or particularly pronounced for, Japan: rapid economic growth, overlapping epidemiological transitions, dietary patterns, traditional health culture, genetic factors, distribution of income and wealth, and social structure. Data limitations make it difficult to quantify effects of these factors on the life expectancy at the national level and compare them internationally. Thus the objective of this paper is to present some hypotheses about reasons for the rapid longevity extension in Japan. Substantiation of those hypotheses is considered future tasks. It should be noted that factors that lengthen the lifespan (level factors) and factors that expedite the lifespan increase (pace factors) are not necessarily the same. For example, gene distributions in a population should affect the life expectancy, but usually the gene distributions in a human population do not change much in a few decades, or even in a few centuries. However, there may be some interactions between level factors and pace factors. For example, in some societies, only a small group of most privileged people benefit from new, expensive medical technologies, whereas in more egalitarian societies, the majority of people may gain access to those medical technologies. In this case, social structure, a relatively stable factor, and medical technology, which may change rapidly, interact in affecting the life expectancy. Furthermore, if a factor is closely related to a certain type of diseases, the importance of the factor for overall mortality does not remain constant but varies with the proportion of deaths from those diseases. For example, a level factor contributing to prevention of chronic diseases may become important as major medical causes of death shifts from acute infectious diseases to chronic diseases, thereby accelerating the overall mortality decline accompanying the change of cause-of-death structure. Thus, this paper discusses both level factors and pace factors without strictly distinguishing them.