This paper explores the factors influencing saving behavior among retired single individuals, focusing on the role of medical expenses. The authors construct a comprehensive model that accounts for bequest motives, heterogeneity in medical expenses and life expectancies, and social insurance programs. Using the Assets and Health Dynamics of the Oldest Old (AHEAD) dataset, they estimate the model's parameters through the method of simulated moments. The results indicate that out-of-pocket medical expenses increase rapidly with age and permanent income, significantly affecting savings decisions. For many elderly, the risk of long-term care and expensive medical expenses is a more significant driver of saving than bequests. Social insurance programs like Medicaid help explain low asset holdings among the poorest, but they also benefit the rich by insuring against the financial risks of very old age. The model's fit is assessed using over-identifying restrictions, and the distribution of deceased individuals' estates generated by the model aligns well with observed data. The study concludes that properly accounting for medical care expenses and social insurance programs is crucial for understanding elderly savings behavior.This paper explores the factors influencing saving behavior among retired single individuals, focusing on the role of medical expenses. The authors construct a comprehensive model that accounts for bequest motives, heterogeneity in medical expenses and life expectancies, and social insurance programs. Using the Assets and Health Dynamics of the Oldest Old (AHEAD) dataset, they estimate the model's parameters through the method of simulated moments. The results indicate that out-of-pocket medical expenses increase rapidly with age and permanent income, significantly affecting savings decisions. For many elderly, the risk of long-term care and expensive medical expenses is a more significant driver of saving than bequests. Social insurance programs like Medicaid help explain low asset holdings among the poorest, but they also benefit the rich by insuring against the financial risks of very old age. The model's fit is assessed using over-identifying restrictions, and the distribution of deceased individuals' estates generated by the model aligns well with observed data. The study concludes that properly accounting for medical care expenses and social insurance programs is crucial for understanding elderly savings behavior.