Forty years after the publication of a landmark paper by the Ad Hoc Committee of the Harvard Medical School, the general concept of brain death has achieved widespread acceptance. In the United States, irreversible dysfunction of the brain and brainstem are required for the diagnosis of brain death. Although primarily based on clinical evaluation, confirmatory examinations, including radionuclide blood flow studies, play an important role in augmenting the physical examination in special situations when some of its specific components cannot be performed or reliably evaluated. The 2 main radionuclidic techniques used in evaluation of brain death are radionuclide angiography with nonlipophilic radiopharmaceuticals and parenchymal imaging with lipophilic agents. Specific technical guidelines for determination of brain death have been promulgated by professional medical societies. In the vast majority of cases, blood flow examinations are useful in confirming brain death. Nonetheless, on occasion patients clinically diagnosed with brain death will exhibit persistent intracranial blood flow or electrical activity. Existence of these contradictory cases reveals underlying inconsistencies in the definitions of brain death. We hypothesize that the existence of these apparent contradictions is related to differences in sensitivity of the physical examination and the confirmatory examinations, differences in localization of the physical examination and confirmatory tests, and differences between blood flow and cerebral function as markers of brain death.