OBJECTIVES: Currently, there are no reliable and validated methods for health service decision-makers to inform policy on quality care at a systems level. To address this need, we worked with health administrators, clinicians and researchers to create and validate a decision-support tool that service planners can use to achieve a sustainable, integrated care system for hip and knee OA. METHODS: The SD model reflects the continuum of care, including self-directed, primary, rheumatologic and orthopaedic specialist for hip and knee replacement, acute, rehabilitation, community and surgical follow-up care. The model was developed in four phases, with phase 1 focusing on demand and flow rates, phase 2 on resources, phase 3 adding geographical stratification, and phase 4 adding feedback loops. We populated the model with administrative data from Alberta Health & Wellness (physician claims, inpatient, and ambulatory data), Survey of Living with Chronic Diseases in Canada, and clinical/surgical data from Alberta Bone and Joint Health Institute. Using established principles of SD modeling and an iterative process of integrated knowledge translation we defined the problem, modeled the system as a series of stock and flows, and validated the model.  Through multiple workshops, experts from front-line clinical staff and administrators provided input and improved face validity. RESULTS: OA care process diagrams were the preferred format for developing the model structure.  The fully specified SD model has been validated with end-users and can be used as a decision-support tool to test scenarios and their resulting effects on system performance and costs. CONCLUSIONS: Based on multiple real-world observational data sources, this SD Model allows evaluation of alternative clinical and administrative scenarios that reflect anticipated changes in health care demands and service. Furthermore, the integrated knowledge translation process reflects the critical importance of involving clinicians and decision-makers when developing a system dynamics model for applied use.