We previously showed that variability of response to clopidogrel is linked to occupancy of the P2Y12 receptor by clopidogrel active-metabolite, and that maximal platelet aggregation intensity (MAI) measured by light transmission aggregometry (LTA) correlates with occupancy. The present study compared a range of ex vivo platelet tests at various levels of P2Y12 occupancy. After screening with clopidogrel 75 mg/day for seven days, subjects were selected to obtain 'low', 'average' and 'high' responders and randomised to clopidogrel (75 mg/day days 1-10; 300 mg day 11), or placebo. Assays were LTA in platelet-rich plasma using 2, 5 and 20 mM ADP, VerifyNow P2Y12, flow cytometric analysis of platelet activation markers and vasodilator-stimulated phosphoprotein (VASP) index, and a thromboelastographic test that is sensitive to clopidogrel. The reference test was P2Y12 receptor occupancy, measured by binding of 33P-2MeS-ADP to platelets. MAI showed the best correlation with P2Y12 occupancy. Similar results were seen with different ADP concentrations and when LTA data were expressed as inhibition of platelet aggregation. A plot of free receptors/cell versus VASP index was biphasic, with poor correlation for low-level P2Y12 occupancy. Sensitivity of the VerifyNow P2Y12 assay decreased at higher clopidogrel responses. Thromboelastography and P-selectin expression had poor correlation with receptor occupancy. In conclusion, LTA data correlate best with P2Y12 occupancy, the gold standard for detecting clopidogrel's effect at the receptor level. Our results highlight a limitation of the VASP index, which appears unable to distinguish low, average and high responders early after clopidogrel initiation when P2Y12 occupancy is still low.