Objectives: To determine the influence of medical expenses on life expectancy. Methods: The expenses of 1,718 municipalities were divided into total expenses, hospitalization expenses and expenses other than hospitalization and dental expenses. 1) The correlation of life expectancy with sex was considered. 2) The correlation between expenses and life expectancy was considered. 3) The correlation of life expectancy or expenses with the numbers of doctors, dentists, facilities and beds was considered. 4) Using the Mahalanobis-Taguchi method, a unit space was formed by 10 municipalities with a high life expectancy, and D2 was calculated. When D2 was outside the unit space, the expenses were not as much as those of the 10 municipalities with a high life expectancy. Results: 1) Life expectancy showed a positive correlation with gender. 2) Male life expectancy showed a negative correlation with total and hospitalization expenses, and a positive correlation with dental expenses. A positive correlation was found between each of expenses and female life expectancy. Total expenses, hospitalization expenses and expenses other than those on hospitalization showed a negative correlations with life expectancy in Hokkaido. Dental expenses showed a negative correlation with life expectancy in Chubu, hospitalization expenses showed a negative correlation with life expectancy in Kyushu. Total, hospitalization and dental expenses showed positive correlations with life expectancy in Tohoku, and dental expenses showed a positive correlation with life expectancy in Kanto and Chubu. 3) Total expenses, hospitalization expenses and expenses other than those on hospitalization were found to correlate with the number of doctors. Dental expenses were found to correlate with the numbers of doctors, facilities, and beds. 4) The difference in among estranged municipalities was considered. Life expectancy was significantly short in estranged municipalities, and the total expenses and hospitalization expenses were large. Conclusions: The relationship of medical expenses with life expectancy became clear. It was assumed that medical performance was poor in estranged municipalities.
