This study explores the preferences and willingness-to-pay of home-based disabled elderly individuals for long-term care insurance (LTCI) compensation mechanisms in Guangzhou, China, using a discrete choice experiment. The research aims to identify preferred compensation strategies, analyze heterogeneity in preferences, and provide recommendations for policy optimization. Using purposive and cluster sampling, 156 eligible participants were identified, with 96 completing the survey (response rate: 61.5.%). Disabled elderly individuals were defined based on activities of daily living assessments. A conditional Logit model was applied to analyze preferences, and subgroup analyses examined differences by education, gender, activities of daily living status, and caregiving arrangements. Key findings include preferences for medical care over life care, family caregivers over professional ones, and cash subsidies over mixed or proportional reimbursement. Respondents were willing to pay an additional $21.60 for medical care and $25.26 for cash subsidies (1 USD = 7.3 CNY). The average out-of-pocket cost for LTCI services was $27.39 per session, with a sub-average cost of $16.44 for basic care services. Subgroup analyses revealed higher-educated individuals favored medical care, while lower-educated groups prioritized affordability. Severely disabled individuals preferred professional caregivers, such as registered nurses. This study highlights the need to expand medical care services, integrate flexible compensation models, and tailor policies to demographic differences. The findings provide evidence for optimizing China's LTCI system and offer insights for aging populations in low- and middle-income countries.