With the deepening of new urbanization and population aging，elderly migrants have become an important component of China’s floating population.Segmenting and identifying the characteristics of the elderly mobile population can help the government provide them with better social security and facilitate their adaptation to the new urban environment.Based on the data of the 2018 national mobile population dynamic monitoring survey，this paper uses the LCA model to segment and identify subgroups within the elderly migrant population.On this basis，the multiple Logistic regression model is further structured to analyze key driving factors behind the mobility of different types of elderly population to provide a basis for social security recommendations.According to the results，the elderly mobile population can be categorized into three types：1） Intergenerational exchange type.Characterized by being older，rural，in poor health，and with generally low household income levels， this group is the most vulnerable among the three，and their migration is largely based on mutual support between generations.2） Economically driven type.Typically younger，rural，in good health，and with multiple children，this group demonstrates the strongest willingness to work or start businesses and migrates primarily for employment or commercial opportunities.3）Emotionally dependent type.Mostly urban，with higher levels of education and household income，this group shows the strongest desire for emotional fulfillment and often migrates with their children for elder care and companionship.This paper suggests that the social security of the elderly mobile population should be enhanced by ensuring the employment participation for elderly mobile population of the “economically driven” group，strengthening social support for the intergenerational exchange type，and optimizing elder care services for the emotionally dependent group.