The purpose of this study was to explore approaches to more accurately assess Caco-2 permeability of poorly water-soluble new chemical entities (NCEs) in an effort to determine their biopharmaceutics classification system (BCS) permeability class with a higher level of confidence. The transport of reference compounds and NCEs (Sch-Y, Sch 56592) was studied across Caco-2 monolayers in the absence or presence of varying percentage of bovine serum albumin (BSA) in the receiver chamber. The inclusion of 0.5-4% BSA in the receiver chamber caused a 4-5-fold increase in Sch-Y P(app), while Sch 56592 P(app) was not significantly influenced. Amongst reference solutes, the P(app) ratio (+BSA/ctrl) was significant (1.3-fold) only for diltiazem (log PC=2.7, plasma protein binding=78%), but the prediction of human oral absorption for such drugs was not affected by the presence of BSA in receiver. In summary, the use of 4% BSA in the receiver chamber during transport studies can dramatically affect the estimated Caco-2 P(app) and BCS permeability ranking of highly lipophilic NCEs, as in the case of Sch-Y with a log PC of 4.0. For Sch-Y, this is presumably due to improved sink conditions and/or a reduction in non-specific drug adsorption to plastic wells. In contrast, the permeability classification of Sch 56592 (log PC=2.4) based on estimated Caco-2 P(app) values is not affected by the presence of receiver BSA.