This research extends the study of social network analysis into the context of long-term care. Network density, reciprocity, and intensity were hypothesized to explain duration of the ties between frail elderly persons and their networks after they enter a residential care home (RCH), which is a type of long-term care facility. Using longitudinal data from interviews with 81 new, elderly RCH residents and 75 of their closest others, multiple regression analyses showed that the density of frail elderly people's networks has the strongest effect on tie duration. Secondary direct effects were also shown for reciprocity, mental status, being state-financed, White, and having returned home. The intensity of elderly people's ties does not explain tie duration.