Penetration of epithelial cells represents the rate-determining step for the absorption of many drugs and pharmaceutical macromolecules such as proteins and nucleic acid therapeutics. While the potential of using cell-penetrating peptides (CPPs) to facilitate absorption has been increasingly recognized, the mechanism of cell penetration and the uptake into certain cells have recently been called into question due to methodological artifacts. Therefore, the objective of this study was to quantitatively assess the ability of RI-Tat-9, a proteolytically stable CPP, to penetrate epithelial cell monolayers. The permeability of RI-Tat-9 with two epithelial cell lines, Madin-Darby canine kidney (MDCK) and Caco-2 cells, was comparable to the leakiness of the respective intact monolayers. Microscopic imaging showed that fluorescence-tagged RI-Tat-9 did not enter these cells, further supporting a paracellular transport mechanism. Although insufficient data were generated in these studies to generalize the observed phenomenon, the entry of RI-Tat-9 into nonepithelial T lymphocytic MT2 cells, possibly by endocytosis, suggested that a cell type-specific barrier might exist that controlled uptake of RI-Tat-9 by cells. Compared to that in MT2 and HeLa cells, the active uptake of the peptide into MDCK monolayers was much slower and showed no dependence of cell energy. Furthermore, the equilibrium binding of RI-Tat-9 to MDCK cells at 0 degrees C was indicative of an interaction with a nonspecific receptor. A correlation between binding density and concentration difference across a leaky separation barrier suggested that repulsion of free peptide molecules by bound peptide molecules at the MDCK monolayer surface may be significant at micromolar concentrations. The results of this study quantitatively show that Tat CPP uptake into two commonly used epithelial cell types is minimal and possibly cell type-specific. Implications for Tat CPP-assisted drug delivery are discussed.