In increasingly constrained health and aged care services, strategies are needed to improve quality and translate evidence into practice. In dementia care, recent failures in quality and safety have led the World Health Organisation to prioritise the translation of known evidence into practice. While quality improvement collaboratives have been widely used in healthcare, there are few examples in dementia care. We describe a recent quality improvement collaborative to improve dementia care across Australia and assess the implementation outcomes of acceptability and feasibility of this strategy to translate known evi-dence into practice. A realist-informed process evaluation was used to analyse how, why and under what circumstances a quality improvement collaborative built knowledge and skills in clinicians working in dementia care. This realist-informed process evaluation developed, tested, and refined the program theory of a quality improvement collaborative. Data were collected pre-and post-intervention using surveys and interviews with participants (n=24). A combined inductive and deductive data analysis process integrated three frameworks to examine the context and mechanisms of knowledge and skill building in participant clinicians. A refined program theory showed how and why clinicians built knowledge and skills in quality improvement in dementia care. Seven mechanisms were identified: motivation, accountability, identity, collective learning, credibility, and reflective practice. Each of these mechanisms operated differently according to context. A quality improvement collaborative designed for clinicians in different contexts and roles was acceptable and feasible in building knowledge and skills of clinicians to improve dementia care. A supportive setting and a credible, flexible, and collaborative process optimises quality improvement knowledge and skills in clinicians working with people with dementia.