Background: Procedural sedation for magnetic resonance imaging (MRI) is frequently required in children to achieve cooperation and immobility for high-quality imaging. Many institutions use propofol infusions alone or in combination with other sedative drugs such as midazolam for sedation during MRI. There are wide ranges of propofol infusion rates reported in the literature (3-18mg/kg/hr) but few details have been reported about the effects of adjunctive midazolam on the dosing of propofol. The practice of our intensivist-led pediatric sedation team for children undergoing MRI typically includes 0.025-0.1mg/kg intravenous (IV) midazolam for anxiolysis followed by 1-2mg/kg propofol boluses and a propofol infusion starting at 5mg/kg/hr, although several patients received propofol alone. We report our ten-year experience of deep sedation for children undergoing MRI to assess the effect of midazolam on propofol dosing. Methods: After IRB approval, records of patients who required procedural sedation for MRI between 2007-2016 were retrospectively reviewed. A total of 2354 patients aged 6 months to 18 years received sedation for MRI at our institution during this period. Drug administration doses were separated into induction (bolus doses in mg/kg given to achieve deep sedation pre-MRI) and total dose (sum of all propofol delivered during MRI and induction boluses expressed as mg/kg/hr since MRI procedures are time variable). Mean doses of propofol for both induction and total were compared to mean doses when IV midazolam was given as an adjunct to the propofol induction. Data was stratified into four age groups and was analyzed using SAS 9.4. Results: 2354 pediatric patients (ASA I&II) underwent sedation for MRI during the 10-year period. Of the patients, 8.5% (n=201) received propofol alone (P) and 75.7% (n=1781) received propofol and IV midazolam (P+M). Excluded patients received a combination of additional adjunct medications (oral midazolam, IV morphine, ketamine, and/or fentanyl). The mean propofol infusion duration for all patients was 46 minutes and the successful MRI completion rate was 99.3%. Additional results are displayed in the table below. For all ages, the mean induction dose of propofol to achieve deep sedation in P+M group was lower than in the P group. In age < 8 years, there was no statistical difference in total propofol dose (mg/kg/hr) between the P+M group and the P group. In the 8-18yr age, the mean total propofol dose was lower in the P+M group (p=0.004). Our data confirmed previous studies that report young patients require higher induction and total propofol doses and that sedation was successful with relatively low infusion rates. Conclusion: For procedural MRI sedation, the addition of anxiolytic doses of IV midazolam lowered the deep sedative induction dose of propofol in all age groups but lowered the total propofol dose only in older children.