Introduction: Early magnetic resonance imaging (MRI) provides important information for management and prognosis in patients with severe traumatic brain injury (sTBI). Yet, optimal timing of MRI remains unknown. The aim of our study was to evaluate the safety of early MRI and to identify a method for appropriate patient selection to minimize adverse events related to the intrahospital transport (IHT) and the MRI examination. Methods: Twenty-six patients with sTBI [mean Glasgow Coma Scale (GCS) 6, range 3-8] admitted to our neurosurgical ICU from 03/2015 to 12/2017 and receiving at least one MRI within the first 14 days after initial traumatic event were prospectively included in the study. The following requirements were fulfilled for at least 4 h prior to anticipated MRI: MAP > 70 mmHg, aPCO 2 30-40 mmHg, stable ICP < 25 mmHg. All relevant cardiopulmonary and cerebral parameters and medication were recorded. The following MRI sequences were performed: DWI, FLAIR, 3D T2-space, 3D T1 MPRAGE, 3D SWI, 3D TOF, pASL, and 1 H/ 31 P-MRS. Results: Four females and 22 males (aged 23-78 years, mean 46.4 years) with a median GCS on admission of 5 (range 3-8) were analyzed. In total, 40 IHTs were performed within the first 14 days (mean 6 days, range 1-14 days). Mean pre-MRI ICP was 14.1 mmHg (range 3-32 mmHg). The mean post-MRI ICP was 14.3 mmHg (range 3-29 mmHg), decreasing to a mean ICP of 13.2 mmHg after 1 h (range 3-29 mmHg). There were no significant differences in ICP measurements before and after MRI (p = 0.30). MAP remained stable with no significant changes during the entire IHT and MRI. No other adverse events were observed as well. Conclusion: Early MRI in acute severe TBI is feasible and safe. Yet, careful patient selection with prior adequate testing of cardiopulmonary and cerebral parameters is crucial to minimize transport-or examination-related morbidity.