Recombinant activated factor VII (rFVIIa) is widely used as a haemostatic agent in non-haemophiliacs with intractable haemorrhage. This practice is based on published case reports and case series describing its haemostatic effect in a variety of clinical settings (Mittal & Watson, 2006). Despite its increasing use, particularly in trauma-related and obstetric haemorrhage (Ahonen & Jokela, 2005; Martinowitz & Michaelson, 2005), evidence provided by randomised controlled trails for the efficacy of rFVIIa is limited to intracerebral haemorrhage (Mayer et al, 2005) and blunt trauma (Boffard et al, 2005). Increasing awareness has led to increasing demand for a resource that remains expensive. Despite analysis of moderate numbers of treated patients, effective tools to predict response and therefore to guide appropriate use have not been developed (Dutton et al, 2004). Between April and June 2006, a survey was sent by electronic mail to haematologists at all 241 hospitals within the UK. A follow-up survey was sent by post or electronic mail to all non-responders in December 2006. Participants were asked to report current rFVIIa prescribing practice within their institution. Responses were obtained from 166 of 241 (69%) of hospitals. A total of 146 of 166 responders (88%) had used rFVIIa in the management of intractable haemorrhage. A total of 139 of 166 (84%) held a stock of rFVIIa for use in this setting, a further 14 (8%) and three (2%) obtained rFVIIa in an emergency from a nearby hospital or direct from the supplier, respectively. Most rFVIIa was issued from the hospital blood bank (75%). Other stock was held in the pharmacy (6%), haemophilia centre (12%), intensive care unit (6%), delivery suite (1%) and operating theatre (1%). Three per cent of hospitals held a supply of rFVIIa in more than one location. In 116 of 156 prescribing hospitals (74%), all rFVIIa use required authorisation by a consultant haematologist. In a further 29 (19%), following a decision to treat by the responsible clinician, each case required discussion with a haematologist prior to administration of rFVIIa. In 11 (7%) of hospitals, rFVIIa was administered without haematological advice, by independent prescribers including anaesthetists, intensivists, obstetricians and surgeons. Prescribing practices varied (Table I). Initial dose ranged from 20 to 400 μg/kg, most frequently 90 μg/kg in 109 of 156 prescribers (70%). Five institutions did not dose by weight, giving a standard dose of 2·4–7·2 mg. Frequency of dosing ranged from every 20 min to 6-h but was most commonly 2-h (31%). One institution reported continuous infusion of rFVIIa over a 24-h period. Five hospitals reported the use of increasing doses with repeated administration. The maximum number of doses was limited to two and three in 38% and 16% of institutions, respectively, whilst 11% stated that further discussion with the haematologist was required in order to administer more than two doses. Total number of uses was reported as 0 in 13%, 0–5 in 49%, 6–20 in 27%, 21–50 in 6% and >50 in 3%. Use of rFVIIa was according to a local or regional protocol in 102 of 156 (65%) institutions and subject to audit in 88 (56%). We conclude that rFVIIa use in the management of intractable haemorrhage is widespread among UK hospitals although dosing regimens vary. A cost of £3700 for a single adult dose [90 μg/kg NovoSeven®, (Novo Nordisk, Crawley, UK) in a 70 kg patient] could be considered difficult to justify in light of unlicensed use of a drug that has little evidence to support its efficacy. A further concern is the risk of thromboembolic events and the lack of safety data when used outside the setting of haemophilia. Authorisation of use remains the responsibility of the haematologist in most UK hospitals; prescribing without the approval of a haematologist occurred in only a few institutions. This provides potential for the development of agreed protocols in order to promote standardised use according to available efficacy and safety data. The authors wish to acknowledge all of the haematologists who responded to this survey and to thank those who additionally provided thoughtful comments.