Background: Among urban and rural elderly in China, comprehensive contextual factors of subjective well-being (SWB) have not yet been investigated systematically. Using a large and representative sample of Chinese elderly people, this study explored the predictors of SWB. As outlined in Andersen’s behavioral model, predictors can be classified as predisposing, enabling, needing, and health factors. Methods: The Chinese General Social Survey (CGSS) is a nationwide, community-based and cross-sectional study conducted between 2003 and 2022. This study used data from 2018. There were 4415 participants aged over 60, including 1763 urban residents and 2652 rural residents. The SWB was measured on a five-point Likert scale ranging from very unhappy to very happy. Estimating potential predictors on SWB was accomplished by using ordinal logistic regression models in the generalized linear method. Results: In China, the SWB of urban and rural elderly has greatly improved, and the gap between them has narrowed significantly. Those with spouses, pensions, higher incomes, better self-reported health, and more social activities have better SWB. In the urban elderly, SWB is primarily determined by enabling factors such as personal income, while in the rural elderly, it is primarily determined by need factors such as social participation. Conclusions: A variety of factors influence SWB among the elderly. An emphasis should be placed on helping widowed elderly people, those without pension insurance, those with low incomes, and those less involved in social activities. Leisure activities, financial support, social security, and medical services can all be improved to promote the SWB of older people.