BACKGROUND Surgical, traumatic and obstetric bleeding are important causes of mortality and morbidity. The antifibrinolytic drug, tranexamic acid (TXA), inhibits clot breakdown and may have a role in the management of excessive bleeding. AIM: Evaluate the evidence from randomised trials for the effects of TXA in patients with surgical bleeding, traumatic bleeding, and for preventing postpartum haemorrhage (PPH). METHODS: Using systematic reviews and meta-analytic techniques, evaluate the evidence from randomised trials to: i) quantify the effects of TXA for surgical bleeding; ii) investigate the quality of trials of TXA for surgical bleeding; iii) to quantify the effects of TXA for traumatic bleeding; iv) estimate the number of avoidable trauma deaths by the routine use of TXA; v) quantify the effects of TXA for preventing PPH; and vi) propose a trial of TXA for preventing PPH. RESULTS: A systematic review including 129 trials involving 10,488 patients suggests that TXA reduces bleeding in surgical patients by about one third, although its effect on death and thromboembolic events is uncertain. Evidence that TXA reduces surgical bleeding has been available for many years, although, poor methodological quality of trials may mean that it is unreliable. There is reliable evidence that TXA reduces death due to bleeding in trauma patients. This evidence originates from a large, high quality randomised trial in 20,211 patients. If TXA was given to all patients soon (<3 hours) after injury over 100,000 deaths could be prevented every year. A systematic review including 26 trials involving 4191 women suggests that there is no reliable evidence for the effects of TXA for preventing PPH. The trials are poor quality and contain serious flaws. The proposed WOMAN-2 trial of TXA for preventing PPH in 10,000 women with anaemia aims to resolve the uncertainties. CONCLUSIONS: Most trials assessing the effect of TXA for surgical bleeding and for preventing PPH are small and poor quality. Although together they provide promising evidence that TXA reduces bleeding, further evidence from large trials at low risk of bias is required to determine reliably the effects of TXA for these indications. There is reliable evidence that TXA reduces the risk of death in trauma patients and no further trials are required. Instead, dissemination of the evidence and implementation of TXA into trauma protocols worldwide, should be a priority. Although there is no evidence from randomised trials that it increases risk, the effect of TXA on thromboembolic events remains uncertain.