1. An understanding of the role that transporters, in particular P-glycoprotein (P-gp), can play in the absorption, distribution, metabolism and excretion (ADME) of candidate drugs, and an assessment of how these processes might impact on toxicity and the potential for drug-drug interactions in the clinic, is required to support drug development and registration. It is therefore necessary to validate preclinical assays for the in vitro evaluation of candidate drugs as substrates or inhibitors of human P-gp. 2. The present study has characterized a Caco-2 cell monolayer model by determining the bi-directional apparent permeabilities and efflux ratios of the known P-gp substrates ([(3)H]-digoxin, [(3)H]-ketoconazole, [(3)H]-verapamil, [(3)H]-quinidine, dipyridamole and loratidine; 1-100 microM) a non-substrate ([(3)H]-propranolol; 10 microM), or by determining the inhibitory potencies (IC(50)) of inhibitors (verapamil, ketoconazole, quinidine, dipyridamole and probenecid; 0.1-100 microM) on the basolateral-to-apical transport of [(3)H]-digoxin (5 microM), in order to validate methodologies for the identification of substrates or inhibitors of P-gp, respectively. 3. The reproducibility of the [(3)H]-digoxin or verapamil data determined from replicate monolayers across different cell passages indicates that the functional expression of P-gp is consistent across the range of passages (25-40) utilized for transport experiments and that the determination of bi-directional apparent permeability, or IC(50) for inhibition of P-gp, respectively, need only be performed on one occasion for a test compound. [(3)H]-digoxin and [(3)H]-propranolol or verapamil and probenecid were considered to be appropriate positive and negative controls of P-gp-mediated transport, or inhibition of P-gp, respectively, to ensure performance of the assays when assessing candidate drugs. Additionally, the low IC(50) values determined for ketoconazole and quinidine indicated that these inhibitors were suitable to use to confirm the role of P-gp in the efflux of a test compound. 4. These validated Caco-2 assays are robust, reproducible and suitable for routine in vitro evaluation of candidate drugs. They have been successfully applied to development projects resulting in the identification of two candidate drugs as substrates and inhibitors of P-gp, whereas a third was neither a substrate nor an inhibitor of this transporter.