Transcranial doppler (TCD) ultrasonography is often used to guide the management of patients with subarachnoid hemorrhage (SAH). However, the correlation between increased blood velocity as measured by TCD ultrasonography and angiographic vasospasm was established before the routine use of hypervolemia/hemodilution and administration of nimodipine and did not address blood flow. The relationship of blood velocity as measured by TCD ultrasonography and local cerebral blood flow (LCBF) in SAH managed with these modalities is unknown. Patients presenting with aneurysmal SAH between January 1992 and September 1993 who underwent TCD ultrasonography and xenon computed tomographic (Xe/CT) LCBF studies within 12 hours were retrospectively studied. Fifty patients underwent a total of 94 paired studies, encompassing 709 vascular territories. All were treated with nimodipine and hypervolemia/hemodilution. Hematocrit, blood pressure, and partial carbon dioxide pressure were similar at the time of TCD ultrasonography and Xe/CT measurement of LCBF. When LCBF in the middle cerebral artery (MCA) was < or = 31 ml/100 g/min, the corresponding peak systolic velocity measured by TCD ultrasonography was 119 cm/s, whereas those > 31 ml/100 g/min had a velocity of 169 cm/s (P = 0.006). High LCBF was associated with high velocity in all vascular territories, reaching significance in all but the internal carotid artery. At the time of each study, 41 neurological examinations were focal and 53 were nonfocal. The Xe/CT measurement of LCBF in the MCA contralateral to a deficit was significantly less than in territories without corresponding clinical deficits (P = 0.01), whereas peak systolic velocities in the MCA were not significantly different (P = 0.71). Territories with increases in blood velocity in the MCA of > 50 cm/s/24 h did not have statistically different LCBF (P = 0.183). Our results suggest that increased blood velocity revealed by TCD ultrasonography correlates with increased LCBF and not with ischemia. No difference in LCBF was found in territories with and without rapid increases in blood velocity in the MCA. Furthermore, although focal neurological deficits corresponded with decreased contralateral LCBF in the MCA, increased velocity did not correlate with neurological findings. Therapeutic decisions based solely on blood velocity revealed by TCD ultrasonography might be inappropriate and potentially harmful. Xe/CT studies of LCBF are useful in guiding the management of SAH.