OBJECTIVE To assess the clinical utility of antithrombin III (AT-III) substitution in adults with septicaemia in an intensive care unit. METHODS A retrospective follow-up study was performed in the adult intensive care unit of a large teaching hospital. Adults with septicaemia and AT-III levels less than 0.45 IU/ml were identified. AT-III administration, consisting of an intravenous bolus injection of 20 IU/kg, followed by continuous infusion of 20 IU/kg per 24 h, was given to 21 patients, while this was withheld in 21 age- and sex-matched controls. The severity of diffuse intravascular coagulation (DIC), APACHE II score, and type of septicaemia were analysed. The odds ratio was calculated for survival. RESULTS The base-line characteristics with regards to severity of DIC, APACHE scores and types of sepsis were comparable for the patients who received AT-III concentrates and those who did not. Mortality in the treated and non-treated groups was 76% (95% CI: 53-92%) and 57% (95% CI: 34-78%), respectively (p = 0.24). The odds ratio for survival was 2.4 if no AT-III concentrate was administered (95% CI: 0.537-11.5; p = 0.24). CONCLUSIONS The use of AT-III concentrates in the doses applied in adult intensive care patients with septicaemia does not appear to improve outcome in terms of mortality.