In the era of aging in China, in order to accurately provide age-appropriate services and design products for the elderly, this paper conducts a study on user portraits of the elderly in the community. The group carried out five ergonomics measurements of the elderly in Guangzhou City, established a data set containing high-dimensional ergonomics characteristics data of 841 elderly people, retained the complete measurements of 210 elderly people’s data including 6 aspects with a total of 12 items, after the principal component analysis (PCA) to reduce the dimension. Two-step clustering algorithm was then used to analyze the total of 6 dimensions (Age, Sex, Hand Strength, Mental State, Cognitive Level, Behavioral Activities) of characteristics of the variables, to constructing six categories of user portraits of elderly people in South China, which show differences in physiology, psychology, cognitive ability and behavioral activities, the study can pro-vide a basis for accurate promotion of products and services for the elderly.