BackgroundLoneliness affects up to 42% of long-term care residents and is associated with poor health outcomes. Humanoid robot interventions hold promise for reducing loneliness and decreasing barriers to social interaction in long-term care settings, such as the current COVID-19 safety measures in many countries, limited mobility, and poor health. We present a protocol describing an assessor-blinded randomized controlled trial comparing the effects of a humanoid robot intervention to treatment as usual, on loneliness and mental health outcomes in long-term care residents.MethodsSeventy-four (n = 74) older adults experiencing loneliness in 3 long-term care homes will be randomized 1:1 to an 8-week, twice a week social intervention with the Grace humanoid robot vs. a treatment as usual active control. We will assess change (baseline to week 8) in (1) loneliness (primary outcome), (2) depression severity, and (3) stress (secondary outcomes), as well as (4) other exploratory outcomes: anxiety, quality of life and reduction in acute healthcare utilization. We will also assess the feasibility and acceptability of the intervention using qualitative methods.DiscussionThe proposed study will assess the effects of a social robot on loneliness and other mental health outcomes, as well as the feasibility of the intervention in older adults living in long-term care settings.Trial registrationNCT05423899.