Background and Aims: Pay-for-performance incentives are used by funders to improve hospital performance and reduce variation, but effectiveness when directed at hospitals remains unproven. In 2012, Queensland Health (Australia) introduced a pay-for-performance scheme to improve stroke unit access. We assessed the effects on known disparities in stroke unit access for patients with stroke from rural and regional areas. Methods: We compared changes in stroke unit access and 6-month mortality following stroke between rural/regional vs metropolitan residents before and after pay-for-performance implementation in 2012. We used interrupted time series analysis on linked hospital admission and death registry datasets from 2009-2017. Inclusions: all adults admitted > 1day with primary ICD-10 diagnosis stroke to any Queensland public hospital (n=39). Exclusions: intracranial haemorrhage under neurosurgical care, subsequent stroke admissions. Results: Prior to pay-for-performance introduction in 2012, fewer rural/regional residents were admitted to stroke units (18% vs 52% metropolitan), and this increased significantly more immediately following pay-for-performance compared to metropolitan residents (37% vs 19% metropolitan; difference 19%, 95%CI 6%, 32%). After 2012, access continued to improve for rural/regional residents (3.4%/year) reaching 78% vs 83% metropolitan (difference 5.1%, 95%CI 1.7, 8.5). Mortality was non-significantly greater in rural/regional residents (20.3% vs 18.5% metropolitan; difference 1.8%; 95%CI -1.8, 5.4), and did not change immediately following introduction of pay-for-performance. Post implementation outcomes for mortality showed downward trends relative to pre-implementation but were not statistically different between the groups. Conclusions: Pay-for-performance had greater impact for rural/regional residents and reduced disparities in stroke unit access, with similar non-significant improvement in mortality trends to metropolitan residents.