BACKGROUND: The number of adults across the globe with significant depressive symptoms has grown substantially during the COVID-19 pandemic. The extant literature supports exercise as a potent behaviour that can significantly reduce depressive symptoms in clinical and non-clinical populations. OBJECTIVE: Using a suite of mobile applications, at-home exercise, including high intensity interval training (HIIT) and/or yoga, was completed to reduce depressive symptoms in the general population in the early months of the pandemic. METHODS: A 6-week, parallel, multiarm, pragmatic randomised controlled trial was completed with four groups: (1) HIIT, (2) Yoga, (3) HIIT+yoga, and (4) waitlist control (WLC). Low active, English-speaking, non-retired Canadians aged 18-64 years were included. Depressive symptoms were measured at baseline and weekly following randomisation. RESULTS: A total of 334 participants were randomised to one of four groups. No differences in depressive symptoms were evident at baseline. The results of latent growth modelling showed significant treatment effects in depressive symptoms for each active group compared with the WLC, with small effect sizes (ESs) in the community-based sample of participants. Treatment groups were not significantly different from each other. Effect sizes were very large (eg, week 6 ES range=-2.34 to -2.52) when restricting the analysis only to participants with high depressive symptoms at baseline. CONCLUSIONS: At-home exercise is a potent behaviour to improve mental health in adults during the pandemic, especially in those with increased levels of depressive symptoms. Promotion of at-home exercise may be a global public health target with important personal, social and economic implications as the world emerges scathed by the pandemic. TRIAL REGISTRATION NUMBER: NCT04400279.