This paper examines why the social gradient of life expectancy does not app
ly in Japan when Okinawa is considered. The social gradient thesis links di
fferences in longevity to social rank, with people and populations in highe
r status hierarchical positions having lower mortality and longer life expe
ctancies than those beneath them in the social scale. Japan has been cited
as a major example of this thesis in that Japanese life expectancy improved
dramatically as Japan rose to the top echelon of nations in economic rank
in the late 20th century. Thus it follows that Japan's most affluent and le
ading prefectures should be the major catalysts behind the nation's rise in
life expectancy as well to the number one position in the world. However,
this is not the case as life expectancy in Okinawa, Japan's poorest prefect
ure, exceeds that of Japan as a whole. We find that the social gradient of
life expectancy does not apply at the prefectural level and question its va
lidity for geographical areas. We suggest that healthy lifestyles, especial
ly diet and the social support of family and friends, are more important th
an sense of hierarchy for longevity in Okinawa.