Brain death (BD) concept has been increasingly widely accepted beginning since the late 1950s, but several controversies have appeared when intracranial pathology is localized to the posterior fossa. In the presence of a primary supratentorial brain lesion, a severe forebrain lesion is combined with either the subsequent gradual loss of brainstem function, due to rostrocaudal transtentorial brain herniation. In secondary brain lesions (i.e., cerebral hypoxia), the brainstem is also affected like the forebrain. However, a minority of patients with a primary infratentorial brain lesion (i.e., basilar artery thrombosis or brainstem or cerebellar bleeds) may retain cerebral blood flow and EEG activity. In this article I discuss that if a brainstem lesion does not provoke a massive increase of intracranial pressure there may be no complete cerebral circulatory arrest, explaining the preservation of EEG activity, evoked potentials, and autonomic function. I also discuss the case of Jahi McMath who was declared brain-dead, but ancillary tests, performed 9 months after initial brain insult, showed conservation of intracranial structures, EEG activity, and autonomic reactivity to “Mother Talks” stimulus, rejecting the diagnosis of BD. Jahi McMath’s MRI study demonstrated a huge lesion in the pons. Some authors have argued that in patients with primary brainstem lesions it might be possible to find a in some cases partial recover of consciousness, even fulfilling clinical BD criteria. This was the case in Jahi McMath. Further research and discussion are necessary about the use of ancillary tests in BD diagnosis in primary posterior fossa lesions.