Rivaroxaban, an oral, direct factor Xa inhibitor, has a dual mode of elimination in humans, with two-thirds metabolized by the liver and one-third renally excreted unchanged. P-glycoprotein (P-gp) is known to be involved in the absorption, distribution, and excretion of drugs. To investigate whether rivaroxaban is a substrate of P-gp, the bidirectional flux of rivaroxaban across Caco-2, wild-type, and P-gp-overexpressing LLC-PK1 cells was investigated. Furthermore, the inhibitory effect of rivaroxaban toward P-gp was determined. Rivaroxaban exhibited high permeability and polarized transport across Caco-2 cells. Rivaroxaban was shown to be a substrate for, but not an inhibitor of, P-gp. Of a set of potential P-gp inhibitors, ketoconazole and ritonavir, but not clarithromycin or erythromycin, inhibited P-gp-mediated transport of rivaroxaban, with half-maximal inhibitory concentration values in the range of therapeutic plasma concentrations. These findings are in line with observed area under the plasma concentration-time curve increases in clinical drug-drug interaction studies indicating a possible involvement of P-gp in the distribution and excretion of rivaroxaban. In vivo studies in wild-type and P-gp double-knockout mice demonstrated that the impact of P-gp alone on the pharmacokinetics of rivaroxaban is minor. However, in P-gp double-knockout mice, a slight increase in brain concentrations and decreased excretion into the gastrointestinal tract were observed compared with wild-type mice. These studies also demonstrated that brain penetration of rivaroxaban is fairly low. In addition to P-gp, a further transport protein might be involved in the secretion of rivaroxaban.