Target levels of ex vivo inhibition of platelet aggregation (IPA) induced by adenosine diphosphate (ADP) that produce clinically relevant effects of clopidogrel, a P2Y12 antagonist, are unclear. We examined standard and modified IPA and P2Y12 receptor occupancy as predictors of antithrombotic (% thrombus weight reduction) and bleeding time (BT, fold-increase over control) effects of clopidogrel in rabbit models of carotid artery thrombosis and cuticle bleeding, respectively. Standard and modified IPA with 20 microM ADP were measured in the absence and presence of partial P2Y1 blockade, respectively. Clopidogrel maximally produced standard IPA of 57% +/- 5%, antithrombotic effect of 85% +/- 1%, BT increase of 6.0 +/- 0.4-fold and P2Y12 receptor occupancy of 87% +/- 5%. Surprisingly, a clopidogrel dose that produced a low standard IPA of 17% +/- 4% and P2Y12 receptor occupancy of 39% +/- 5% achieved a significant antithrombotic activity of 55% +/- 2% with a moderate increase in BT of 2.0 +/- 0.1-fold. This underestimation of clopidogrel efficacy by standard IPA was improved by measuring either modified IPA or P2Y12 receptor occupancy. These results suggest that in clopidogrel-treated rabbits, low standard IPA is associated with significant antithrombotic effects. Moreover, modified IPA and P2Y12 receptor occupancy appear to better predict the magnitude of clopidogrel's efficacy compared with standard IPA, which may be a better predictor of BT.