The present study investigates whether changes in total brain blood flow can be reliably estimated by changes in carotid arterial blood flow in fetal and perinatal lambs. We therefore compared carotid arterial blood flow, measured with implanted transit-time ultrasound transducers, with brain blood flow, measured by radioactive microspheres in fetal lambs during normal oxygenation and during pulmonary ventilation with oxygen, with PO2 ranging from levels normal for the healthy fetus to levels normally seen postnatally. Cerebral perfusion pressure was modified over a wide range to alter brain blood flow: it was decreased by balloon occlusion of the brachiocephalic trunk and increased by a balloon occluder around the aortic isthmus. Carotid arterial blood flow and brain blood flow were closely related (r = 0.97, p < 0.0001). The relationship was not altered at different levels of oxygenation. However, measurements during higher cerebral perfusion pressures, obtained during aortic isthmus occlusion, had a negative influence on the agreement between carotid arterial blood flow and brain blood flow. When excluding values obtained by aortic isthmus occlusion, changes of 20% or more in brain blood flow could be predicted with carotid arterial blood flow within a confidence limit of 95%. Blood flow measurements in the carotid artery may be useful to estimate changes in brain perfusion.