This paper deals with differences between European countries in the use of long term elderly care. We analyse whether differences between countries in the use of health care hold, even if we take into account demographic and health differences. In the use of health care we distinguish between informal care (defined as care by family members, other relatives, friends and neighbours) and formal care (defined as paid or professional care).The analysis is based on the SHARE-data: the Survey of Health, Ageing and Retirement in Europe. In the Spring of 2004 about 23.000 persons aged 50 years and older in 11 countries were interviewed. This yielded information on demographics, physical and mental health, and use of health care of about 1.400 households per country. In our analysis 9 countries are included: The Netherlands, France, Germany, Austria, Denmark, Sweden, Spain, Italy and Greece. Cross-tabulations provide indications of country differences in both the use of formal and informal care. Countries with more than avarage informal care are The Netherlands, Greece and Italy. Countries with more than avarage formal care are The Netherlands, France and Denmark. In Greece and Italy more than avarage informal care use coincides with less than avarage formal care use. In France and Denmark it is the other way round: these two countries have al lot of elderly persons receiving formal care and few elderly persons receiving informal care. The Netherlands can be characterized by much informal care as well as much formal care.The starting point of the analysis is a conceptual model, in which three groups of factors of the use of care are distinguished: needs of care, available resources and social status. For needs of care we contructed measures for physical and mental impairments. For available resources we use the availability of an informal network and income. Social status is instrumented by level of education, age, sexe and level of urbanisation.To find out whether the country differences hold, even if we correct for differences in the explanatory factors, we have estimated a nested logit model explaining whether an elderly person receives care, whether it is formal or informal and what type of care it is. The results show that even after correcting for differences between countries in health, available resources and social circumstances, there still is a significant difference in the use of care. An impaired elderly person in one of the Southern countries has a lower probability of receiving care than the same elderly person in one of the Nordic countries. And if an impaired elderly person receives care, the probability of receiving formal care is lower in the Southern countries than in the Nordic countries
