Intravenous recombinant tissue plasminogen activator (rt-PA) is given for acute ischemic stroke using a weight-based dosing regimen, and potential medication dosing errors may impact the relative risks and benefits of this therapy. Weight is frequently estimated by the patient, the family, the nurse, or the treating physician. Discrepancies between actual and estimated weight result in an incorrect dose, but errors of this type have not been previously studied in clinical practice. We hypothesized that such errors may impact the risks and benefits of rt-PA for stroke. The Multicenter rt-PA Acute Stroke Survey included data on 1205 acute stroke patients treated in routine clinical practice with intravenous rt-PA. We calculated the actual unit dose (in mg/kg) by dividing the dose of rt-PA given by the actual weight, and correlated this with risk of intracerebral hemorrhage (ICH) and likelihood of good recovery (modified Rankin score of 0 or 1). Seven hundred and sixty-nine patients (64%) had data on both weight and rt-PA dosage. Forty-one patients (5.4%) had a symptomatic hemorrhage while 51 (6.6%) had an asymptomatic hemorrhage. There were non-significant trends towards increased risk of any ICH as the degree of overdosage increased, particularly in the highest dose quintile compared to the four lower quintiles (15.8% v 11.0%, P = .097). Adjustment for age, baseline NIHSS, and major early computed tomography (CT) changes strengthened this association (16.5% v 9.3%; P = .025). There was no association between actual dose and likelihood of good recovery (P = .57). Overdosage of rt-PA cause by an overestimation of weight resulted in a modest increase in the risk of ICH in the highest quintile, but there did not appear to be any reduction in effectiveness caused by underdosing. Every effort should be made to obtain the most accurate weight.