Background: Financial incentives are an unproven mechanism to improve hospital performance and patient outcomes. In 2012, Queensland Health supported by a state-wide stroke clinical network, introduced a pay-for-performance incentive scheme to improve stroke unit access. Payments (average AUD200,000/year per hospital [n=20]) were contingent on achievement of incremental performance targets, endorsement of stroke unit processes against national guidelines and submission of clinical indicator data. Aims: To assess the effects of the pay-for-performance incentive on: stroke unit access; 6-month mortality; and death or institutional care following acute stroke across a state-wide public health system. Methods: Longitudinal cohort study of stroke and myocardial infarct (MI: control condition) using linked hospital admission and death registry datasets from July 2009 to June 2017 before vs after pay-for-performance implementation in 2012. Inclusions: all adult public hospital admissions >1 day for stroke or MI. Exclusions: intracranial haemorrhage admitted under neurosurgical care, subsequent events within study period. Interrupted time series analysis was used to assess change of level and trends in outcomes; with historical trends and MI as controls. Results: Pay-for-performance introduction in 2012 was associated with an increase in stroke unit access of 30.9% (95%CI 23.8, 38.1) above baseline trends, and subsequent reversal of control period non-significant upward trend in 6-month mortality (+1.3%/year) to a significant decline (-0.6%/year); difference -1.9%/year; (95%CI -3.5, -0.4). In contrast for MI, there was no change in mortality which continued non-significant downward trends (-0.5%/year before, -0.2%/year after 2012). A 2.2%/year greater change in stroke mortality following pay-for-performance introduction compared to MI was found (difference-in-difference, 95% CI -4.1, -0.2). Death or residential care following initial hospitalisation followed similar, but non-significant trends. Conclusions: Pay-for-performance, implemented through a clinical network, stimulated significant improvements in access to acute stroke unit care; and a sustained decline in 6-month mortality compared to both historical and MI controls.