The drug transporter P-glycoprotein (P-gp) influences drug distribution across the blood-brain barrier (BBB) by actively extruding drugs into the neural capillaries. Several psychotropic drugs, including nortriptyline (NT) and risperidone (Risp), are substrates of P-gp. Here we compared the in vitro P-gp interactions of Risp and its major metabolite, 9-OH-Risperidone (OH-Risp), with their distribution over the BBB in P-gp knock-out mice and in rats where P-gp was inhibited. K(m) and V(max) were determined by an in vitro ATPase assay, and V(max)/K(m) ratios of 2.7 and 0.5 were recorded for Risp and OH-Risp, respectively, suggesting that Risp is a better substrate for P-gp than OH-Risp. In Mdr1a (-/-) knock-out mice, the brain-serum ratios of both Risp and OH-Risp were more than ten times those of control mice (14 and 11, respectively). When P-gp was inhibited with cyclosporine A (CsA) in Wistar rats, the effect was an order of magnitude less than that observed for the knock-out mice experiments (1-1.5 times the controls), and co-administration of NT had no effect. In conclusion, both Risp and OH-Risp interact with P-gp in vitro, and P-gp has a profound effect on Risp and OH-Risp distribution over the BBB, as is evident from the knock-out mice experiments. Drug-drug interaction effects in relation to P-gp, however, appear to be more limited.