The present work was undertaken to examine the central pharmacokinetics of phenytoin (PHT) in an experimental model of epilepsy, induced by administration of 3-mercaptopropionic acid (MP), and possible participation of P-glycoprotein in this model of epilepsy. Repeated seizures were induced in male Wistar rats by injection of 3-MP (45mgkg −1 , i.p.) during 10 days. Control rats (C) were injected with saline solution. In order to monitor extracellular PHT levels, either a shunt microdialysis probe or a concentric probe was inserted into carotid artery or hippocampus, respectively. All animals were administered with PHT (30mgkg −1 , i.v.) 30min after intraperitoneal administration of vehicle (V) or nimodipine (NIMO, 2mgkg −1 ). No differences were found in PHT plasma levels comparing all experimental groups. In pre-treated rats with V, hippocampal PHT concentrations were lower in MP (maximal concentration, C max : 2.7±0.3μgml −1 , p<0.05 versus C rats) than in C animals (C max : 5.3±0.9μgml −1 ). Control rats pre-treated with NIMO showed similar results (C max : 4.5±0.8μgml −1 ) than those pre-treated with V. NIMO pre-treatment of MP rats showed higher PHT concentrations (C max : 6.8±1.0μgml −1 , p<0.05) when compared with V pre-treated MP group. Our results indicate that central pharmacokinetics of PHT is altered in MP epileptic rats. The effect of NIMO on hippocampal concentrations of PHT suggests that P-glycoprotein has a role in reduced central bioavailability of PHT in our epileptic refractory model.