BACKGROUND AND PURPOSE: We sought to evaluate predictors of clinical outcome, angiographic success, and adverse effects after intra-arterial administration of urokinase for acute ischemic stroke. METHODS: We designed a Brain Attack program at University Hospitals of Cleveland for diagnosis and treatment of patients presenting within 6 hours of onset of neurological deficit. Patients with ischemia referable to the carotid circulation were treated with intra-arterial urokinase. Angiographic recanalization was assessed at the end of medication infusion. Intracerebral hemorrhage was investigated immediately after and 24 hours after treatment. Stroke severity was determined, followed by long-term outcome. RESULTS: Fifty-four patients were treated. There was improvement of >/=4 points on the National Institutes of Health Stroke Scale from presentation to 24 hours after onset in 43% of the treated patients, and this was related to the severity of the initial deficit. Forty-eight percent of patients had a Barthel Index score of 95 to 100 at 90 days, and total mortality was 24%. Cranial CT scans revealed intracerebral hemorrhage in 17% of patients in the first 24 hours, and these patients had more severe deficits at presentation. Eighty-seven percent of patients received intravenous heparin after thrombolysis, and 9% of them developed a hemorrhage into infarction. Angiographic recanalization was the rule in complete occlusions of the horizontal portion of the middle cerebral artery, but distal carotid occlusions responded less well to thrombolysis. CONCLUSIONS: The intra-arterial route for thrombolysis allows for greater diagnostic precision and achievement of a higher concentration of the thrombolytic agent in the vicinity of the clot. Disadvantages of this therapy lie in the cost and delay. Severity of stroke and site of angiographic occlusion may be important predictors of successful treatment.