BACKGROUND: Decades of research indicate that when social connectedness is threatened, mental health is at risk. However, extant interventions to tackle loneliness have had only modest success, and none have been trialled under conditions of such threat. METHOD: 174 young people with depression and loneliness were randomised to one of two evidence-based treatments: cognitive behaviour therapy (CBT) or Groups 4 Health (G4H), an intervention designed to increase social group belonging. Depression, loneliness, and well-being outcomes were evaluated at one-year follow-up; COVID-19 lockdown restrictions were imposed partway through follow-up assessments. This provided a quasi-experimental test of the utility of each intervention in the presence (lockdown group) and absence (control group) of a threat to social connectedness. RESULTS: At one-year follow-up, participants in lockdown reported significantly poorer wellbeing than controls who completed follow-up before lockdown, t(152)=2.41, p=.017. Although both CBT and G4H led to symptom improvement, the benefits of G4H were more robust following an unanticipated threat to social connectedness for depression (χ(16)=31.35, p=.001), loneliness (χ(8)=21.622, p=.006), and wellbeing (χ(8)=22.938, p=.003). LIMITATIONS: Because the COVID-19 lockdown was unanticipated, this analysis represents an opportunistic use of available data. As a result, we could not measure the specific impact of restrictions on participants, such as reduced income, degree of isolation, or health-related anxieties. CONCLUSIONS: G4H delivered one year prior to COVID-19 lockdown offered greater protection than CBT against relapse of loneliness and depression symptoms. Implications are discussed with a focus on how these benefits might be extended to other life stressors and transitions.