Intra-arterial infusion of urokinase (UK) has been widely used. However, the optimal infusion rate of the reagent has never been determined. This was investigated in the acute stage of middle cerebral artery (MCA) embolism in the present study. Sprague Dawley male rats (n=43) were randomly divided into sham-operation and five ischemic groups with urokinase administration at different infusion rates or without urokinase administration. Ischemia was induced with MCA embolism. Two hours after embolism, total urokinase (urokinase, 170,000U/kg) was given in groups A,B,C and D (n=8 each) at different rates: 1,000 U (0.03 ml/min) per minute, 4,000U (0.12 ml/min), 10,000U (0.30 ml/min), and 16,000U (0.48 ml/min), respectively. Group E received normal saline at a rate of 0.48 ml/min. The sham-operation group (no embolism) received urokinase at (170,000U/kg, 1.5 ml, 16,000 U/min). During ischemia and thrombolysis, regional cerebral blood flow (CBF) was monitored by laser Doppler flowmetry. The neurological deficits, infarct volumes and mortalities in each group were determined. The CBF in ischemic hemisphere were significantly (p<0.05) decreased after embolism in groups A∼E at similar levels (27.32±8.20% to 34.71±6.84%). After different treatments, in group B 4,000U/min infusion of UK induced the best reperfusion, the least neurological deficits and infarct volume, as well as the least mortality and lowest incidence of hemorrhage. The effect of intra-artery thrombolysis of urokinase was related to the infusion rate. Our study demonstrated an optimal infusion rate at 4,000U/min, suggesting relatively low levels of infusion are better able to improve brain reperfusion and reduce brain injury after stroke.