The impact of P-glycoprotein (P-gp) on the distribution of nortriptyline (NT) and its metabolites between brain, liver and serum was studied experimentally. The interaction of NT with P-gp in vitro was confirmed by measurement of P-gp stimulated ATPase activity (Km = 257.6 microM, Vmax = 51.0 nmol phosphate released/mg protein-min). Administration of NT (5 mg/kg, s.c.) to mdrla knockout mice resulted in enhanced brain-serum (1.6-fold, p = 0.012) and liver-serum (1.4-fold, p = 0.019) ratios, as compared to the wild-type mice. For a series of NT doses (2.5, 5, 10, 25, 30 mg/kg, i.p.) inhibition of P-gp with cyclosporine A (CsA, 200 mg/kg, i.p.) in rats led to NT brain- and liver-serum ratios that were on average 1.3- (p = 0.005) and 2.1- (p = 0.001) fold higher than those of the controls, respectively. Verapamil (50 mg/kg) (NT, 5 mg/kg) increased the ratios by a factor of 1.6 (p <0.001) and 10.3 (p <0.001) for brain and liver, respectively. Finally, co-administration of methadone (1 mg/kg) did not alter the brain-serum ratio of NT, but in the liver a slight increase (1.5-fold, p = 0.035) was observed. In conclusion, verapamil yielded complete inhibition of P-gp at the blood-brain barrier and CsA had an effect corresponding to about 50% inhibition. The results show that P-gp influences the penetration of NT into the brain, and that drug-drug interactions may take place.