Objective: This study compared the cost of an integrated primary-secondary care general practitioner (GP)-based Beacon model with usual care at hospital outpatient departments (OPDs) for patients with complex type 2 diabetes. Methods: A costing analysis was completed alongside a non-inferiority randomised control trial. Costs were calculated using information from accounting data and interviews with clinic managers. Two OPDs and three GP-based Beacon practices participated. In the Beacon practices, GPs with a special interest in advanced diabetes care worked with an endocrinologist and diabetes nurse educator to care for referred patients. The main outcome was incremental cost saving per patient course of treatment from a health system perspective. Uncertainty was characterised with probabilistic sensitivity analysis using Monte Carlo simulation. Results: The Beacon model is cost saving: The incremental cost saving per patient was A$365 (95% confidence interval-A$901, A$55) and was cost saving in 93.7% of simulations. The key contributors to the variance in the cost saving per patient course of treatment were the mean number of patients seen per site and the number of additional presentations per course of treatment associated with the Beacon model. Conclusions: Beacon clinics were less costly per patient course of treatment than usual care in hospital OPDs for equivalent clinical outcomes. Local contractual arrangements and potential variation in the operational cost structure are of significant consideration in determining the cost-efficiency of Beacon models. What is known about this topic?: Despite the growing importance of achieving care quality within constrained budgets, there are few costing studies comparing clinically-equivalent hospital and community-based care models. What does this paper add?: Costing analyses comparing hospital-based to GP-based health services require considerable effort and are complex. We show that GP-based Beacon clinics for patients with complex chronic disease can be less costly per patient course of treatment than usual care offered in hospital OPDs. What are the implications for practitioners?: In addition to improving access and convenience for patients, transferring care from hospital to the community can reduce health system costs.