OBJECTIVE: To evaluate the impact of the Australian National Emergency Access Target (NEAT) policy introduced in 2012 on ED performance. METHODS: A longitudinal cohort study of NEAT implementation using linked data, for 12 EDs across New South Wales (NSW), Australian Capital Territory (ACT) and Queensland (QLD) between 2008 and 2013. Segmented regression in a multi-level model was used to analyse ED performance over time before and after NEAT introduction. The main outcomes measures were ED length of stay ≤4 h, access block, number of ED presentations, short-stay admission (≤24 h), >24 h admissions, unplanned ED re-attendances within 7 days and 'left at own risk' (including 'did not wait for assessment'). RESULTS: Two years after NEAT introduction, ED length of stay ≤4 h increased in NSW and QLD (odds ratio [OR] = 2.48 and 3.24; P < 0.001) and access block decreased (OR = 0.41 and 0.22; P < 0.001), but not in ACT (OR = 1.28; P > 0.05). ED presentations increased over time before and after NEAT introduction with a significant increase above the projected trend in NSW after NEAT (mean ratio = 1.07). Short-stay admissions increased in QLD (OR = 2.60), ACT (OR = 1.68) and NSW (OR = 1.35). Unplanned ED re-attendances did not change significantly. Those who left at their own risk decreased significantly in NSW and QLD (OR = 0.38 and 0.67). CONCLUSION: ED presentations continued to increase over time in all jurisdictions. NSW and QLD, but not ACT, showed significant improvements in time-based measures. Significant increases in short-stay admissions suggest a strategic change in ED process associated with NEAT implementation. Rates of unplanned ED re-attendances and those leaving at their own risk showed no evidence for adverse effects from NEAT.