Neonatal stroke remains a complex pathophysiologic process that is poorly understood and difficult to investigate, The primary animal model used to study this phenomenon is that of unilateral carotid artery ligation with 2-3 hours exposure to severe hypoxia, A new model of neonatal stroke was developed based on transient middle cerebral artery occlusion without craniectomy, In this model a #6-0 (0.07 mm) nylon filament is passed via the carotid artery to occlude reversibly the middle cerebral artery for 4 hours under conditions of normoxia in 14- to 18-day-old spontaneously hypertensive rat pups, After removal of the filament and reperfusion for 24 hours, the infarct volume was determined using the mitochondrial stain, 2,3,5-triphenyltetrazolium chloride, Using this technique, a neocortical and caudoputamenal infarct affecting 49% of hemispheric volume that measured 180 +/- 29 mm(3) (hemisphere volume = 359 +/- 16 mm(3), mean +/- SEM) was created in 90% of animals (n = 8) undergoing this procedure, This model has the advantage of being relatively noninvasive, of not requiring global exposure of brain to hypoxia, and of using temporary rather than permanent occlusion, This technique should improve the ability to study the acute and long-term pathophysiology of neonatal stroke, particularly the phenomenon of reperfusion injury, as well as its sequelae in the developing nervous system.