Renal failure causes multiple physiological changes involving CNS dysfunction. In cases of uremia, there is close correlation between plasma levels of uremic toxins [e.g. 3-carboxy-4-methyl-5-propyl-2-furanpropionate (CMPF), hippurate (HA) and indoleacetate (IA)] and the degree of uremic encephalopathy, suggesting that uremic toxins are involved in uremic encephalopathy. In order to evaluate the relevance of uremic toxins to CNS dysfunction, we investigated directional transport of uremic toxins across the blood-brain barrier (BBB) using in vivo integration plot analysis and the brain efflux index method. We observed saturable efflux transport of [(3)H]CMPF, [(14)C]HA and [(3)H]IA, which was inhibited by probenecid. For all uremic toxins evaluated, apparent efflux clearance across the BBB was greater than apparent influx clearance, suggesting that these toxins are predominantly transported from the brain to blood across the BBB. Saturable efflux transport of [(3)H]CMPF, [(14)C]HA and [(3)H]IA was completely inhibited by benzylpenicillin, which is a substrate of rat organic anion transporter 3 (rOat3). Taurocholate and digoxin, which are common substrates of rat organic anion transporting polypeptide (rOatp), partially inhibited the efflux of [(3)H]CMPF. Transport experiments using a Xenopus laevis oocyte expression system revealed that CMPF, HA and IA are substrates of rOat3, and that CMPF (but not HA or IA) is a substrate of rOap2. These results suggest that rOat3 mediates brain-to-blood transport of uremic toxins, and that rOatp2 is involved in efflux of CMPF. Thus, conditions typical of uremia can cause inhibition of brain-to-blood transport involving rOat3 and/or rOatp2, leading to accumulation of endogenous metabolites and drugs in the brain.