High on-treatment platelet reactivity (HTPR) after a clopidogrel loading dose predicts the risk of thrombotic events after percutaneous coronary intervention. We have demonstrated that HTPR could be overcome in most cases using dose adjustment according to PR monitoring resulting in an improved clinical outcome. However, this strategy failed in nearly 10% of patients with HTPR. Cytochrome P450 (CYP) 2C19 polymorphism was a major determinant of the response to clopidogrel and could be responsible for a failure of dose adjustment. We aimed to determine the clinical and genetical predictors of a failure of the dose-adjustment strategy. Seventy-three patients undergoing percutaneous coronary intervention were included in this prospective multicenter study. A vasodilator phosphoprotein index >or=50% after a 600-mg loading dose of clopidogrel defined HTPR. Dose adjustment was performed according to PR monitoring to reach a vasodilator phosphoprotein index <50%. Genetic polymorphism of CYP2C19 was determined by direct sequencing. Clinical predictors of HTPR were body mass index (BMI; p = 0.01), diabetes mellitus (p = 0.03), and acute coronary syndrome (p = 0.02). The mutant 2 allele of CYP2C19 681A > G loss of function polymorphism was also significantly associated with HTPR (p = 0.04). The rate of successful dose adjustment was similar in carriers of the CYP2C19 2 allele and carriers of the wild-type allele. The only independent predictor of a failed dose adjustment was a high BMI (p = 0.01). In conclusion, high BMI, acute coronary syndrome, diabetes mellitus, and CYP2C19 2 are associated with HTPR after a 600-mg loading dose of clopidogrel. Dose adjustment overcomes HTPR in carriers of the CYP2C19 2 allele. BMI is the only independent predictor of failed dose adjustment. Thus, drug underdosage seems to be the main determinant of HTPR.