ABSTRACT: In animal models of ischaemic stroke, 25% albumin reduced brain infarction and improved neurobehavioural outcome. In a pilot clinical trial, albumin doses as high as 2 g/kg were safely tolerated. We aimed to assess whether albumin given within 5 h of the onset of acute ischaemic stroke increased the proportion of patients with a favourable outcome.We did a randomised, double-blind, parallel-group, phase 3, placebo-controlled trial between Feb 27, 2009, and Sept 10, 2012, at 69 sites in the USA, 13 sites in Canada, two sites in Finland, and five sites in Israel. Patients aged 18-83 years with ischaemic (ie, non-haemorrhagic) stroke with a baseline National Institutes of Health stroke scale (NIHSS) score of 6 or more who could be treated within 5 h of onset were randomly assigned (1:1), via a central web-based randomisation process with a biased coin minimisation approach, to receive 25% albumin (2 g [8 mL] per kg; maximum dose 750 mL) or the equivalent volume of isotonic saline. All study personnel and participants were masked to the identity of the study drug. The primary endpoint was favourable outcome, defined as either a modified Rankin scale score of 0 or 1, or an NIHSS score of 0 or 1, or both, at 90 days. Analysis was by intention to treat. Thrombolytic therapies were permitted. This trial is registered with ClinicalTrials.gov, number NCT00235495.422 participants were randomly assigned to receive albumin and 419 to receive saline. On Sept 12, 2012, the trial was stopped early for futility (n=841). The primary outcome did not differ between patients in the albumin group and those in the saline group (186 [44%] vs 185 [44%]; risk ratio 0·96, 95% CI 0·84-1·10, adjusted for baseline NIHSS score and thrombolysis stratum). Mild-to-moderate pulmonary oedema was more common in patients given albumin than in those given saline (54 [13%] of 412 vs 5 [1%] of 412 patients); symptomatic intracranial haemorrhage within 24 h was also more common in patients in the albumin group than in the placebo group (17 [4%] of 415 vs 7 [2%] of 414 patients). Although the rate of favourable outcome in patients given albumin remained consistent at 44-45% over the course of the trial, the cumulative rate of favourable outcome in patients given saline rose steadily from 31% to 44%.Our findings show no clinical benefit of 25% albumin in patients with ischaemic stroke; however, they should not discourage further efforts to identify effective strategies to protect the ischaemic brain, especially because of preclinical literature showing convincing proof-of-principle for the possibility of this outcome.National Institute of Neurological Disorders and Stroke, US National Institutes of Health; and Baxter Healthcare Corporation.