Brain death is a clinical diagnosis and must be confirmed by one paraclinical test. This report presents the usefulness of paraclinical tests applied during a 4-year period. Forty-four patients with severe brain lesions leading to brain death were treated during the 2004-2007 period. The appropriate test was chosen according to test availability and patient condition, appreciating restrictions according to the test protocol. Since the results of some tests were inconclusive, some patients underwent repeat testing with the same or different methods. Among 44 patients, 19 had neurotrauma, 11 massive aneurysmal subarachnoid hemorrhage, 1 massive arteriovenous subarachnoid and parenchymal hemorrhage, 12 hypertensive parenchymal hemorrhage, and 1 ischemic stroke. As a primary test, transcranial Doppler (TCD) was used in 30, brain scintigraphy in 2, multislice computed tomography angiography (MSCTA) in 10, and cerebral angiography in 2 patients; the diagnosis was confirmed in 26, 3, 9 and 2 patients, respectively. Due to inconclusive test results, MSCTA had to be repeated in 4 patients twice, and in one patient three times. Four patients where TCD was used died during the observation period, and in one patient the hemodynamic spectrum was inconclusive. In most patients (65%), TCD confirmed the clinical diagnosis of brain death, and in 61% the diagnosis was confirmed within a 2-hour period. TCD was the most useful confirmatory test for cerebral circulatory arrest in brain death diagnosis.