Context.-Rivaroxaban is a new oral anticoagulant that functions as a direct anti-Xa inhibitor. Although routine monitoring is not required, measurement of plasma concentrations may be necessary in certain clinical situations. Routine coagulation assays, such as the prothrombin time and, to a lesser degree, activated partial thromboplastin time, correlate with drug concentration, but because of reagent variability, these methods are not reliable for determining rivaroxaban anticoagulation.Objective.-To compare different methods and calibrators for measuring rivaroxaban, including the chromogenic anti-Xa assay, which, when calibrated with a rivaroxaban standard, may be more appropriate for determining anticoagulation.Design.-We compared measured rivaroxaban concentrations with the same anti-Xa kit but used different calibrators, with different anti-Xa kits but the same calibrators, with antithrombin-supplemented anti-Xa kit versus nonsupplemented kits, and with 2 methods based on rivaroxaban-calibrated, high-phospholipid, dilute Russell viper venom time. Regression and paired t test statistics were used to determine correlation and significant differences among methods and calibrator sources.Results.-Although there was strong correlation, statistically significant biases existed among methods that report rivaroxaban levels. A single-source calibrator did not alleviate those differences among methods. High-phospholipid Russell viper venom reagents correlated with rivaroxaban concentration but were not better than chromogenic anti-Xa methods.Conclusions.-Rivaroxaban-calibrated, anti-Xa measurements correlate well, but the clinical significance of the variation with rivaroxaban measurements is uncertain. The antithrombin-supplemented, anti-Xa method should be avoided for measuring rivaroxaban.