Background Adrenaline is recommended during cardio-pulmonary resuscitation. The optimal dose remains debated, and the effect of lower than recommended dose is unknown. Objective To compare the outcome of patients treated with the recommended, lower or higher cumulative doses of adrenaline. Design, settings, participants Patients were included from the French National Cardiac Arrest Registry and were grouped based on the received dose of adrenaline: recommended, higher and lower dose. Outcome measures and analysis The primary endpoint was good neurologic outcome at 30 days post-OHCA, defined by a cerebral performance category (CPC) of less than 3. Secondary endpoints included return of spontaneous circulation and survival to hospital discharge. A multiple propensity score adjustment approach was performed. Main results 27 309 patients included from July 1st 2011 to January 1st 2019 were analysed, mean age was 68 (57–78) years and 11.2% had ventricular fibrillation. 588 (2.2%) patients survived with a good CPC score. After adjustment, patients in the high dose group had a significant lower rate of good neurologic outcome (OR, 0.6; 95% CI, 0.5–0.7). There was no significant difference for the primary endpoint in the lower dose group (OR, 0.8; 95% CI, 0.7–1.1). There was a lower rate of survival to hospital discharge in the high-dose group vs. standard group (OR, 0.5; 95% CI, 0.5–0.6). Conclusion The use of lower doses of adrenaline was not associated with a significant difference on survival good neurologic outcomes at D30. But a higher dose of adrenaline was associated with a lower rate of survival with good neurological outcomes and poorer survival at D30.