BACKGROUND: We considered whether haemorrhage after a liver trauma would be reduced by early administration of a pro-haemostatic agent and evaluated the effect of i.v. vs i.m. administration of the coagulation factor VIIa analogue NN1731 on haemorrhage after a liver trauma in the pig. METHODS: The pharmacokinetics of i.v. and i.m. NN1731 was evaluated in eight minipigs, and the effects of dose and administration route of NN1731 (i.v. 180 microg kg(-1), n=6; i.m. 540 microg kg(-1), n=4, or 2000 microg kg(-1), n=6) vs vehicle (n=16) were studied on a liver laceration injury in pigs. To simulate a pre-hospital setting, the administration of NN1731 was delayed by 1 min for i.m. administration and 7 min for i.v. administration, at which time fluid resuscitation also began. RESULTS: In the minipigs, NN1731 exposure was similar after i.v. 180 microg kg(-1) and i.m. 540 microg kg(-1), with a bioavailability of approximately 35%. The injury and blood loss at 7 min was comparable between the four groups of pigs; however, after 60 min, the blood loss was lower in the i.v. treated animals: 1.3 (0.3) (i.v.) vs 2.2 (0.8) litres (i.m.(540), i.m.(2000), and vehicle) (P<0.001). Also, the survival time was increased: 117 (14) (i.v.) vs 84 (28) min (i.m.(540), i.m.(2000), and vehicle) (P<0.001). CONCLUSIONS: After a liver trauma in the pig, i.v. administration of NN1731 reduced the bleeding and increased the survival time. In contrast, i.m. administration had no effect, presumably because reduced muscle perfusion during haemorrhage reduced the uptake of NN1731.