Objectives Limited social connectedness in older adults is a risk factor for poor physical and mental health. Older adults who are socially isolated, lonely and disconnected have a higher risk of chronic illness, depression and premature death. Current literature suggests that improved social connectedness reduces these risks. Intergenerational programmes are an effective way to improve health outcomes. Despite this, there is yet to be a review using realist review methods that seeks to identify the circumstances that promote social connectedness in older adults participating in intergenerational programmes with adolescents. Design A realist review methodology was chosen to account for the complexity of intergenerational interventions. Nine studies were included. In line with realist review methodology, iterative data extraction and analysis was conducted to identify the specific contexts, mechanisms and outcomes of the programmes. Specific circumstances were identified to develop theories relating to improved social connectedness in older adults. Data sources MEDLINE, PsycINFO, CINAHL were searched using English language limitation. Eligibility criteria Included participants were aged 65 and over (older adults) and between 13 and 19 years (adolescents) participating in intergenerational programmes from non-familial generations. Studies had to be published in English between 2000 and 2020 and could be quantitative, qualitative or mixed-methods primary research studies. Data extraction and synthesis Two independent reviewers used a bespoke data extraction form. All authors were involved in the synthesis process which used the extracted data to illuminate the contexts, mechanisms and outcomes that underpinned reviewed programmes. Results The nine included studies were set in different contexts, including community organisations, schools and aged care facilities. They used an array of interventions including reminiscence therapy, craft or space for conversation. Despite study heterogeneity, the parallels in psychosocial development between older adults and adolescents were shown to be a likely driver for improved social health outcomes. Programmes most likely to improve social health outcomes were those that acknowledged psychosocial development, were delivered in community settings, leveraged pedagogical frameworks, used trained facilitators and supported participants to build relationships through shared purpose. Conclusions This review contributes a logic model to support the design and development of intergenerational programmes involving adolescents to improve social connectedness in older adults. Future research to test the logic model in practice is needed.