Abstract Antithrombotic agents are effective but pose significant bleeding risk, thus safer new agents are needed. Since the infused double mutant human thrombin (W215A/E217A; WE) binds endothelial thrombomodulin, activates protein C locally in the intraluminal boundary layer, and prevents thrombus formation without hemostasis impairment, we investigated whether WE could also interrupt the growth of existing thrombi. WE was compared to standard interventional doses of the low molecular weight heparin, enoxaparin (Sanofi) in awake baboons. Thrombosis was initiated by interposing a two part device consisting of a 2 cm, 4 mm ID polyethylene terephthalate vascular graft (DVG) followed by a 2 cm silicone rubber expansion chamber, 9 mmID. Upon initiation of blood flow at 100 mL/min, wall shear rate of 265 sec-1 in the 4 mm ID segment (arterial type flow) and 29 sec-1 in the expansion chamber (venous type flow), platelet thrombus was monitored by gamma camera imaging of autologous 111In-labelled platelets for 60 min or 90 min. Fibrin accumulation was quantified by homologous 125I-labelled fibrinogen. Hemostasis was assessed as template bleeding time prolongation. APTT was also monitored. Treatments started 25 min after the initiation of thrombosis by enoxaparin, 0.5 mg/kg (LD) and 1 mg/kg (HD) IV bolus or WE, 7.5 μg/kg/hr (LD) and 15 μg/kg/hr (HD) infusion (1/3 dose given as a bolus followed by IV infusion over 60min) and compared to saline infused controls. All interventions were given systemically, downstream from the device and 5–6 studies comprised each study group. Treatment with enoxaparin at 25 min after initiation of thrombus growth resulted in a reduction of platelet deposition of 19 % with the LD and 76 % with the HD after 35 min of therapy in the expanded venous type segment. WE reduced venous thrombosis by 45 % and 65 % at the two doses studied and was further reduced to 65 % and 80 % following 65 min of WE infusion as compared to controls. Arterial type thrombosis was reduced by WE by 34 % and 39 % respectively for the two doses studied. HD enoxaparin reduced arterial type thrombosis by 37 %. Fibrin accumulation paralleled the platelet deposition. APTT was prolonged 1.57- and 1.87-fold respectively for the WE doses without any bleeding time prolongation. Enoxaparin prolonged the APTT 1.4- and 1.75-fold, similar to WE but with a 1.13- and 1.75-fold prolongation of the bleeding time. In conclusion, the double mutant human thrombin, WE, effectively interrupted both venous and arterial type growing thrombi at very low doses without detectable compromise to hemostasis. Likewise, enoxaparin at the doses tested had similar efficacy but with a significantly prolonged bleeding time at the 1 mg/kg bolus dose. WE might provide a safer alternative to heparins in the treatment of acute progressive thrombosis.