Tranexamic acid, an antifibrinolytic agent, reduces postoperative transfusion requirements but carries a poorly documented risk of thrombosis. Does tranexamic acid reduce mortality among victims of severe traumatic bleeding? To answer this question, we conducted a review of the literature, using the standard Prescrire methodology. The Crash-2 trial compared the impact of tranexamic acid versus placebo on overall mortality in 20 211 trauma victims with either severe bleeding or a high risk of severe bleeding. Overall mortality at 4 weeks was 14.5% in the tranexamic acid group versus 16% in the placebo group (p = 0.004); on average, one death was prevented when about 67 patients were treated with tranexamic acid. A statistically significant reduction in bleeding-related mortality was observed when tranexamic acid was given within 1 hour after injury. The benefit appeared to decline over time, and when treatment began more than 3 hours after injury, it appeared to increase the risk of death due to bleeding. In this trial, the effect of tranexamic acid on overall mortality among patients with traumatic head injury is uncertain. Rapid intravenous injection of tranexamic acid can cause hypotension. Cases of thrombosis have also been reported. In 2011, the FDA warned of a risk of convulsions linked to high intravenous doses of this drug. In practice, intravenous tranexamic acid infusion has a favourable harm-benefit balance in patients with severe traumatic bleeding, especially when treatment begins less than 1 hour after injury, in which case it reduces mortality at 4 weeks.