Objectives: To determine the clinical and economic impact of a community-based, hybrid model of in-person and virtual care by comparing health-system performance of the rural jurisdiction where this model was implemented with neighbouring jurisdictions without such a model and the broader regional health system. Design: A cross-sectional comparative study. Setting: Ontario, Canada, with a focus on three largely rural public health units from April 1, 2018, until March 31, 2021. Participants: All residents of Ontario, Canada under the age of 105 eligible for the Ontario Health Insurance Plan (OHIP) during the study period. Interventions: An innovative, community-based, hybrid model of in-person and virtual care, the Virtual Triage and Assessment Centre (VTAC), was implemented in Renfrew County, Ontario on March 27, 2020. Main outcome measures: Primary outcome was change in emergency department (ED) visits anywhere in Ontario, secondary outcomes included changes in hospitalizations and health-system costs, using percent changes in mean monthly values of linked health-system administrative data for two years pre-implementation and one year post-implementation. Results: Renfrew County saw larger declines in ED visits (-34.4%, 95% confidence interval -41.9% to -26.0%) and hospitalizations (-11.1%, 95% confidence interval -19.7% to -1.5%), and slower growth in health-system costs than other rural regions studied. VTAC patients' low-acuity ED visits decreased by -32.9%, high-acuity visits increased by 8.2%, and hospitalizations increased by 30.0%. Conclusion: After implementing VTAC, Renfrew County saw reduced ED visits and hospitalizations and slower health-system cost growth compared to neighbouring rural jurisdictions. VTAC patients experienced reduced unnecessary ED visits and increased appropriate care. Community-based, hybrid models of in-person and virtual care may reduce the burden on emergency and hospital services in rural, remote and underserved regions. Further study is required to evaluate potential for scale and spread. Trial registration: Not applicable.