The purpose of this study is to examine on breast shapes of elderly women and provide fundamental data for developing brassiere. Numerous anthropometric measurements and other related data from 398 subjects were analyzed by various statistical methods such as ANOVA, Correlation analysis, Factor analysis and Cluster analysis. 1. Analysis of Breast Measurements The subjects were categorized into three groups(Group I; aged 55 to 64, Group Ⅱ; aged 65 to 74, Group Ⅲ; aged more than 75). 1) The results of the comparative analysis of anthropometric data from three groups show that by getting older: ①Breast lengths, widths, heights and bust girths are decreased significantly. ②Various length measurements related to the drooping degree of breast are increased. ③Some measurements related to the volume of breast are decreased. 2) The results of the correlation analysis among measurement show that there are no direct linear relationships between under bust girth and bust drooping. Further it turned out that the cup size could be used as a factor explaining the volume of breast due to large subject variation, Thus it is required to have more specific information about the breast volume. 2. Analysis of Breast Shapes 1) From 17 measurements, 5 factor were selected as key factors for the factor analysis of breast analysis of breast shapes. The 5 factors are: ①Drooping degree of breast ②slope between breast and chest, width of bust point ③Contours and prominence of breast at the point of front and side ④Breast volume ⑤Breast width. 2) We categorized the breast shapes into three types by Cluster analysis. Type 1 is the most common breast shape in elderly women. (①Type 1; Not too droopy and large breast ②Type 2; A little droopy and small breast ③Type 3; Very droopy and wide breast