Traumatic brain injury (TBI) occurs abruptly, involves multiple specialized teams, calls on the health-care system in its emergency dimension, and engages the well-being of the patient and his relatives for a lifetime period. Clinicians in charge of these patients are faced with issues of uppermost importance: medical issues such as predicting the long-term neurological outcome of the comatose patient; ethical issues because of the influence of intensive care on the long-term survival of patients in a vegetative and minimally conscious state; legal issues because of the law that has set the concept of proportionality of care as the legal rule; and social issues as the result of the very high cost of these pathologies. Today's larger availability of magnetic resonance imaging (MRI) in ventilated patients and the recent improvements in hardware and in imaging techniques that have made the last-developed imaging techniques such as diffusion tensor imaging and magnetic resonance spectroscopy available in brain-trauma patients, are changing the paradigm in neurointensive care regarding outcome prediction. The old paradigm that no individual prognosis could be made at the subacute phase in TBI patients does not hold true anymore. This major change opens new challenging ethical questions. This review focuses on the brain explorations that are required, such as MRI, magnetic resonance spectroscopy, and diffusion tensor imaging, to provide the clinician with a multimodal assessment of the brain state to predict outcome of coma. Such an assessment will become mandatory in the near future to answer the crucial question of proportionality of care in these patients.