Study objectives: To review the randomized, controlled, multicenter trials of intravenous thrombolytic therapy for ischemic stroke. Methods: Studies of ischemic stroke confirmed by computed tomography (CT) and randomization of more than 100 patients are reviewed. Streptokinase studies are the MAST-I, the MAST-E, and the ASK Trial. Studies using tissue plasminogen activator (tPA) are the NINDS Stroke Study, ECASS I, ECASS II, and ATLANTIS. One study using ancrod is STAT. We discuss significant factors common to each study, including thrombolytic agent used, CT scan interpretation, time of therapy administration in relation to stroke onset, thrombolytic dose, ancillary medication administration, safety, and neurologic outcomes. Results: All streptokinase studies were stopped early because of increased mortality in the treated groups. Initial results of the STAT study are promising; publication of full study details is awaited. The ATLANTIS study was terminated early because of nonstatistical efficacy at interim analysis. The NINDS and the ECASS trials were completed; only the NINDS study demonstrated significant increase in the percentage of patients with complete recovery or minimal deficit at 3 months, without significant difference in mortality in the treated group. Conclusion: This review supports the use of intravenous thrombolytic therapy for ischemic stroke using tPA at a dose of .9 mg/kg body weight and a 'golden window' treatment time of 3 hours. Administration without strict adherence to protocol, even within this time frame, may shift the benefit/risk profile of tPA. We recommend the treating physician have rapid access to CT scanning and to collaboration with individuals experienced in the evaluation of stroke and CT interpretation.