The effect of delayed thrombolysis with recombinant tissue plasminogen activator was tested in an embolic stroke model. The carotid territory was embolized in 103 rats with fibrin-rich clots formed and washed in polyethylene tubes. Hemispheric cerebral blood flow before and after embolization was measured by the intra arterial 133Xe injection method. At five delay times, 15-240 min after embolization, 69 animals were treated with tissue plasminogen activator, 20 mg/kg, and 34 animals with saline. Carotid angiography displayed the grade of occlusion of the cerebral arterial supply before and after treatment. Brains were fixed after 2 days, evaluated neuropathologically, and infarct volume measured. Cerebral blood flow was reduced by 56-71% after embolization. Reperfusion induced by thrombolytic therapy was demonstrated by comparing the posttreatment angiography of the pooled five treatment groups to control animals. Thrombolytic therapy significantly reduced the infarct volume and improved the prekill clinical score by up to 2 h of treatment delay, and treatment might have been beneficial even after 4 h delay. Prolonging the delay of treatment increased the infarct volume (p < 0.001, Jonck-heere-Terpstra test). Only a few hemorrhagic complications were observed. Thus, thrombolytic therapy in embolic stroke induced recanalization. The effect on clinical outcome and infarct volume was dependent on delay time.