Although a partner is generally considered the major provider of support and company (Cantor 1991) and living close to family is associated with the frequency of contacts (Bordone 2009, Hank 2007) and exchange of support (Knijn & Liefbroer 2006, Mulder & van der Meer 2009), the presence of partners and the geographic proximity of children might of course proxy also other factors such as the underl. [...] More specifically, geographic characteristics such as population size, the relative number of elderly, the general health of the local population and the focus on home health versus institutionalized care may bear some relevance for the provision of LTC, net of individual factors and perhaps also net of the characteristics of family members. [...] Elderly persons with neither a partner nor a child who is disadvantaged have the lowest risks of transitioning to LTC, irrespective of the geographical proximity of the child, and the risk is about 50 percent lower than that of unpartnered, childless elderly. [...] A comparison between the upper and lower panels in Figure 3 shows that the differences in general are greater for the disadvantaged than for the advantaged composite variable, in line with the results in Figures 1 and 2. [...] Men and women with partners who are not disadvantaged (categories 5-9) have the lowest risk of institutionalization, irrespective of the geographic proximity and/or the resources of the child.