We determined the dose response of the ADP antagonist clopidogrel (0.3-50 mg/kg p.o.) in rat models of thrombosis and provoked bleeding and correlated these activities to ex vivo platelet activation. Carotid artery thrombosis was induced by FeCl(2). Bleeding time was measured by mesenteric vessel puncture and renal cortex or cuticle incision. Platelet biomarkers included standard ADP-induced aggregation, P2Y(12) receptor occupancy, and phosphorylation of vasodilator-stimulated phosphoprotein. Clopidogrel decreased thrombus weight up to 78%, caused maximal prolongation of cuticle and mesenteric bleeding, but had little effect on renal bleeds. Due to the steep mesenteric dose response, further comparisons concentrated on cuticle bleeding. The half-maximal inhibitory dose (ED(50)) for thrombus reduction was 2.4 +/- 0.4 mg/kg, with 10 mg/kg providing optimal blood flow preservation and thrombus reduction. The ED(50) for bleeding was 10.5 +/- 3.4 mg/kg. Increased bleeding was intermediate (3-fold) at 10 mg/kg and maximal (6-fold) at 30 mg/kg. All biomarkers were affected, but with differing sensitivity. ED(50)s for peak platelet aggregation to 10 microM ADP (11.9 +/- 0.4 mg/kg) and the vasodilator-stimulated phosphoprotein index (16.4 +/- 1.3 mg/kg) approximated the higher ED(50) for bleeding. ED(50)s for ligand binding (3.0 +/- 0.3 mg/kg) and late aggregation (5.1 +/- 0.4 mg/kg) better matched the lower ED(50) for antithrombotic activity. Aspirin exerted lesser effects on bleeding (42-70% increase in all models) and thrombosis (24% inhibition). In summary, antithrombotic doses of clopidogrel have limited effects on bleeding and standard measures of platelet aggregation. Other biomarkers may be more sensitive for tracking antithrombotic efficacy.