Against the backdrop of an aging global population and the increasing pressure of medical care expenditures for seniors, this paper used a fuzzy regression discontinuity (FRD) model to explore the effects of retirement on the self-assessed health and objective physical and mental health of older people. Using survey data from the China Health and Retirement Longitudinal Study (CHARLS), our model addresses some relevant academic controversies. Our sample was comprised of male respondents from government agencies, enterprises, and public institutions. The research explored the impact of retirement on lifestyle habits and included an in-depth analysis of the mechanism through which retirement influences different aspects of health. The results show that: (1) Retirement does not have any significant impact on objective health, including depression and self-care ability, but it does cause a notable decline in subjective health assessment. (2) Retirement shortened the sleep time of respondents, which may account for lower scores on subjective health self-evaluations, but it did not lead to any noticeable improvement in habits which are harmful to health, such as smoking and drinking. (3) Marriage can help alleviate the problems of depression and smoking among older people, and education has a somewhat broader positive effect on their health and lifestyles; however, neither factor helps to improve the sleep problems of older people. Therefore, this paper recommends that efforts should be made to both optimize retirement policies and seek further ways to improve the health of the retired population.