The scale and distribution of the disabled elders are important indicators for estimating demands of the elderly for long-term care services. In addressing the estimation biases caused by inadequate sample size in current elderly surveys, this study pools the survey data of CLHLS, SSAPUR and CHRLS in 2010～2011 weighted by the population structure of the sixth population census, and estimates the disability proportion and the healthy life expectation of the elderly. Results show that the disability rate of the Chinese elderly in 2010 stood at 11.2%, and the average living time duration with ADL Disability was 2.53 years; rural and female elderly were more likely to suffer from moderate and severe ADL disability than their urban and female counterparts. In addition, the gender difference in the proportion of the healthy life expectation among total life expectation rises as age grows. Conclusively, this study suggests that rural and female elderly are the dominant population groups receiving long-term care services.

The magnitude and trend of the disabled elderly population are important basis for evaluating demands of the elderly for long-term care services and planning development of social service system in an aging society. However, domestic existing data and research have not reached consensus on this issue. Using functions status information of the elderly in multiple surveys, this study examines the criteria of disability status in these surveys, and compares the disability rates reported from those surveys, with an attempt to obtain a reasonable estimation of the disability rate of the elderly in China. The analysis of data of three special surveys on ageing shows that KATZ Scale is an comparable tool usually adopted for the measurement of function status in those surveys, and the disability rate of Chinese elderly is between 10.48% and 13.31%. In order to facilitate further study on trends and patterns of disability of the elderly，this paper suggests that a consistent measurement in disability of the elderly for data collection be adopted, the sample size of the oldest-old people be expanded, and the representativeness of samples for different regions be improved in future research surveys.