An improved two-chamber diffusion cell was developed for the study of drug transport across cultured cell monolayers. The cell monolayer was grown on a horizontal support membrane of polycarbonate, which could be rotated providing theoretically predictable thicknesses of the diffusion boundary layer in the donor and the receiver solution as a function of rotation rate. Permeation measurements were performed using the support membrane with and without a cell monolayer. The Madin-Darby-Bovine-Kidney (MDBK) cell line was employed, and permeability coefficients of model solutes (salicylic acid, mannitol, testosterone) across the three distinct mass transport barriers (i.e., the cell monolayer, the support membrane, and the diffusion boundary layer) were determined. The permeability of the diffusion boundary layer followed the theoretical dependence on the rotation rate; absolute values, however, deviated from predictions. Permeability coefficients for all three transport barriers varied substantially between solutes. This variation was the strongest for the permeability coefficient for the cell monolayer and resulted in a varying relative significance of these three barriers in controlling permeation kinetics. This improved diffusion apparatus permits the measurement of unbiased permeability values of solutes across the cell monolayer, notably when the cell monolayer is not absolutely the rate-determining barrier.