Abstract Background In social prescribing, link workers support individuals whose persistent health problems are exacerbated by loneliness by connecting them to community-based social activities. This approach is well established in the UK and is gaining attention in Australia. However, a major limitation of research to date has been a lack of theoretically informed and rigorous evaluations of social prescribing. We will address these points in this study, applying a social identity framework to examine the effects of group-based social prescribing (SP) activity compared to primary care treatment as usual (TAU). Methods Ninety participants experiencing loneliness recruited from primary care services and community centres across five sites in Southeast Queensland will be assigned to one of two conditions (SP, TAU) and assessed at two timepoints (baseline, + 8 weeks). Individuals will be aged 18 years and over, have sufficient English language skills to provide consent, and at the time of recruitment they will not be experiencing acute symptoms or social issues that require urgent intervention. Primary outcomes are loneliness, mental well-being, and health service use (total number of GP, hospital, and allied health visits in the past 3 months). Secondary outcomes will assess social group processes, including number of important social groups, new group identification, multiple identity compatibility, and group-based support and emotion regulation. Discussion This study will provide comprehensive data about the extent to which, and how, social prescribing to community-based group activities may help people to feel less lonely, more socially integrated, and healthy over the first 8 weeks. If effective, this social identity-informed model of social prescribing can be disseminated in communities across Australia. Trial registration ANZCTR, Registered 8 June 2022 - Retrospectively registered, https://www.anzctr.org.au/ACTRN12622000801718.aspx