Introduction: An evidence-based care protocol to treat migraine with low-dose Propofol was implemented in May 2014 at the emergency department (ED) of the CHUL (Québec city). Given potential side effects of Propofol, we aimed to evaluate the safety and effectiveness of this protocol. Methods: We reviewed charts of all patients aged 16 years and older who received Propofol between May 2014 and August 2017 for a migraine headache with or without aura, as defined in the International Headache Society Classification. The care protocol consisted of: 1) administration of intra-venous Propofol 20 mg each 5 to 10 minutes as needed (maximum of 6 doses); 2) sets of vital signs before and after each dose; and 3) continuous cardiac and saturation monitoring. Our primary outcome measures were the incidence (95%CI) of the following side effects: low arterial pressure (< 90 systolic or < 65 mean), desaturation (SaO2<92%), excessive sedation (scores 3 or 4 on the Pasero scale), and any arrhythmia. We also compared the mean reduction (95%CI) of pain pre- and post-treatment (visual analogous scale VAS 0-10) and the proportion (95%CI) of rescue medication among patients who received Propofol as first-line medication to a matched cohort of patients who had Metoclopramide first. The two cohorts were paired for gender, age, triage priority, and month/year of ED visit. Results: Over the 3-year study period, 45 patients with migraine received Propofol through the care protocol, either as a first-line or a rescue therapy. In this cohort, hypotension, bradycardia (<60/min) and desaturation occurred in 17.8% (8.0-32.1), 13.3% (5.1-26.8) and 6.7% (1.4-18.3) of cases respectively; no excessive sedation was reported. An intervention was undertaken in 4 cases [8.9% (2.5-21.2) 3 iv fluid bolus, 1 supplemental oxygen] to palliate the side effects of Propofol. A statistically significant mean reduction of 3.6 points (2.8-4.4) on the VAS scale was observed in patients treated with Propofol as first-line therapy (n=35). However, patients managed with first-line Metoclopramide (n=100) experienced a significantly higher mean reduction of their VAS score [5.3 (4.6-6.0)] than the Propofol group (p=0.003). The proportion of patients requiring the use of rescue medication was higher among patients first treated with Propofol [77.1% (63.2-91.1) vs. 29.0% (20.1-37.9); p<0.001]. Conclusion: Our care protocol to treat migraine with low doses of Propofol appears to be safe and to cause very few side effects prompting corrective interventions. Continuous (as opposed to intermittent) heart and saturation monitoring is probably not indicated. Given the effectiveness of Propofol compared to Metoclopramide, our care protocol will be used as a second-line therapy.