SUMMARYPurpose: The rational discovery and development of new antiepileptic drugs (AEDs) with safer therapeutic index and better pharmacokinetic properties is still warranted nowadays. Because the long-term management of epilepsy is attained by means of orally administered AEDs, investigation of their potential to be well absorbed at the intestinal level is mandatory. Moreover, involvement of the efflux transport mediated by P-glycoprotein (P-gp) may compromise the systemic and central nervous system disposition of AEDs. Therefore, this study aimed at characterizing mouse jejunal passive transport and the possible active efflux mediated by P-gp of a series of dibenz, carbamazepine-10, 11-epoxide[CBZ-E],10,11-trans-dihydroxy-10,11-dihydrocarbamazepine [trans-diol], and BIA 2-024), which comprise some AEDs and metabolites. Methods: Permeation studies were performed with freshly excised mouse jejunum segments mounted in Ussing chambers. Absorptive (M-S) and secretive (S-M) transports were analyzed with and without verapamil, which is a P-gp inhibitor widely recognized. Apparent permeability coefficients (P app ) in both directions and in absence or presence of verapamil were determined for each test compound. The in vitro method was validated using five controls that included high and low permeable markers with known absorption fraction (Fa) and also well-known P-gp substrates. The integrity of intestinal membrane was guaranteed during the assay by measuring the transepithelial electrical resistance. Key Findings: The correlation obtained between P app (M-S) and Fa of references was high (r 2 = 0.9945), and could be used to classify the derivatives according to Biopharmaceutical Classification System: CBZ and OXC were the only classified as highly permeable. The P app (S-M) of OXC, CBZ-E, R-Lic, and BIA 2-024 were significantly higher than their P app (M-S). After verapamil addition, their P app (S-M) lowered while P app (S-M) increased, suggesting the involvement of P-gp on the transport of those compounds across mouse jejunum segments. In opposition, CBZ, S-Lic, and trans-diol presented no statistical differences between the P app values reported in both directions, with or without verapamil. The results reported herein suggest that differences in biodisposition of S-Lic and R-Lic might result from their distinct interaction with P-gp. Significance: The Ussing chamber model used herein showed to be useful for predicting Fa of AEDs and the involvement of efflux transport, namely P-gp, on their absorption. This is an important achievement as compounds that are not transported by P-gp may offer advantages when used in patients with pharmacoresistant epilepsy.