Thrombin has been implicated as a contributing factor to restenosis after vessel reopening procedures. We compared the ability of the direct thrombin inhibitor recombinant (r-) hirudin to reduce neointimal growth in different animal models of arterial injury. Carotid arteries of rats, rabbits, and hypercholesterolemic minipigs were injured by withdrawal of an inflated balloon catheter. In addition, we used a double-lesion model in rabbits, which involved balloon angioplasty of a preexisting lesion induced by carotid denudation 4 weeks earlier. r-Hirudin was given in all four animal models as a short-term application (bolus of 1 mg/kg IV immediately before injury, followed by infusion of 1 mg · kg⁻¹ · h ⁻¹ for 2 hours, and an injection of 6 mg/kg SC). Additionally, we investigated the effects of prolonged treatment (intravenous infusion for 3 and 14 days) in rats. Inhibition of thrombin was monitored by determination of activated partial thromboplastin time, and histomorphometric analysis of the arteries was performed after 2 (rats) or 4 (rabbits and minipigs) weeks. In rabbits, short-term r-hirudin treatment reduced neointimal area by 59% (single-injury model, P=.05) and 44% (double-injury model, P=.02). In rats and minipigs no inhibition of neointimal growth was observed after short-term r-hirudin application. A 3- or 14-day infusion of r-hirudin in rats, however, resulted in 25% (P=.007) and 27% (P=.003) reductions in neointimal area, respectively. In conclusion, there is considerable interspecies variation in the time frame of susceptibility for reduction of neointimal growth by inhibition of thrombin after arterial injury. These results demonstrate the importance of testing potential antirestenotic treatments in an array of different animal models.