Serum thromboxane B2 (TxB2) is a specific marker of platelet inhibition by aspirin. Yet, TxB2 levels differ by up to 10-fold between some aspirin-treated patient cohorts. This study aimed to identify factors responsible for differences in serum TxB2 between cohorts in the ADRIE study (n = 657) and the BOSTON study (n = 678) of aspirin-treated cardiovascular patients originally tested with different ELISA assays. TxB2 levels were assessed in representative subgroups of the two cohorts (34 samples in BOSTON and 39 in ADRIE) by both ELISAs, as well as liquid chromatography and tandem mass spectroscopy (MS). A multivariate analysis was performed on the whole cohort database to identify determinants of the difference of TxB2 levels between cohorts. There was no systematic bias between the original ELISA TxB2 values and the MS values and the median difference was small, 0.12 ng/ml, thus not explaining the difference between median TxB2 levels in the two study populations (7 and 0.6 ng/ml in the ADRIE and BOSTON studies, respectively). In the combined dataset of the ADRIE and BOSTON cohorts (n = 1342), body mass index, age, gender, aspirin dose, time from aspirin intake to blood draw, NSAID intake, platelet count and C-reactive protein were significantly associated with TxB2 levels. After adjustment for patient characteristics, the difference between cohorts did not decrease. Unexplained differences in serum TxB2 levels in different populations of aspirin-treated cardiovascular patients suggest that further studies are needed to confirm the role of serum TxB2 level as a prognostic factor or rather as a marker of therapeutic observance.