Concerns persist regarding the safety of tissue plasminogen activator (tPA) therapy for acute ischemic stroke. Numerous case series of clinical experience with tPA have been published that provide additional data on the safety of thrombolytic therapy. This is a meta-analysis of 15 published, open-label studies that broadly followed approved indications and guidelines for tPA use in nonselective patient populations. In 2639 treated patients, the symptomatic intracerebral hemorrhage rate was 5.2% (95% confidence interval, 4.3 to 6.0), slightly lower than the 6.4% rate in the treated group of the randomized, placebo-controlled National Institute of Neurological Disorders and Stroke (NINDS) trial. The mean total death rate (13.4%) and proportion of subjects achieving a very favorable outcome (37.1%) were comparable to the NINDS trial results. Protocol deviations were reported in 19.8%. Comparing across studies showed that the mortality rate was correlated with the percentage of protocol violations (r=0.67, P=0.018). Postapproval data support the safety of intravenous thrombolytic therapy with tPA for acute ischemic stroke, especially when established treatment guidelines are followed.