Introduction: Magnetic resonance imaging (MRI) is a common procedure in tertiary care neonatal intensive care units (NICUs). MRIs aid in detailing structural anatomy and are increasingly utilized for prognostication. Keeping babies calm and motion-free in the MRI suite is challenging, and various approaches have been adopted to obtain the best image quality. We share our experience of intervention bundle for procedural sedation with the novel use of buccal midazolam in our NICU for babies undergoing MRI. Methods: This single-center quality improvement project comprised two epochs. Epoch 1 from April 2018 to December 2020 provided baseline data regarding sedation use and helped identify causes for suboptimal images and the adverse event rate. Following the implementation of an interventional bundle comprising specific midazolam dose recommendations tailored to background risk factors and streamlining the procedural sedation process, similar comparative data were collected in epoch 2 (May 2021 to December 2022) after a washout period. Results: Of 424 patients, 238 and 108 had MRI done under either procedural sedation protocol or feed and wrap technique in epoch 1 and 2, respectively. After excluding babies whose MRIs were performed under sedative infusions, 30 (13%) babies had adverse events in epoch 1, while only 8 (7%) events occurred in epoch 2. There was also a 37% improvement in the documentation of procedural sedation between the two epochs. Conclusion: Procedural sedation with buccal midazolam under neonatologist supervision is safe, efficient, and effective in babies undergoing MRI in this single-center study. More extensive studies may be warranted to assess the suitability of this sedation modality for broader use.