Loneliness is a common problem in long-term care facilities (LTCF) and previous work has shown that animal-assisted therapy (AAT) can to some degree reverse loneliness. Here, we compared the ability of a living dog (Dog) and a robotic dog (AIBO) to treat loneliness in elderly patients living in LTCF. In comparison with a control group not receiving AAT, both the Dog and AIBO groups had statistically significant improvements in their levels of loneliness. As measured by a modified Lexington Attachment to Pets Scale (MLAPS), residents showed high levels of attachment to both the dog and AIBO. Subscale analysis showed that the AIBO group scored lower than the living dog on "animal rights/animal welfare" but not on "general attachment" or "people substituting." However, MLAPS measures did not correlate with changes in loneliness, showing that attachment was not the mechanism by which AAT decreases loneliness. We conclude that interactive robotic dogs can reduce loneliness in residents of LTCF and that residents become attached to these robots. However, level of attachment does not explain the decrease in loneliness associated with AAT conducted with either a living or robotic dog.