Parallel artificial membrane permeability assay (PAMPA) is a screening tool for the estimation of drug permeability across various biological membrane. We evaluated passive gastrointestinal absorption of fourteen new thiourea derivatives of naproxen using PAMPA test. Diffusion through artificial membranes that were composed of a mixture of hexadecane and hexane (the first PAMPA model) and of the solution of egg lecithin in dodecane (the second PAMPA model) was tested for the set of seven compounds and parent drug naproxen. The starting solutions were prepared by dissolving tested compounds in a phosphate buffer (pH 5.5), with the addition of DMSO (1%) and TWEEN80 (0.2%). These solutions were transferred into the donor PAMPA plates (300 µl) in triplicates. The acceptor solution (phosphate buffer (pH 7.4) and the same percentage content of DMSO and TWEEN as in the starting solutions) was transferred into the acceptor PAMPA plate (300 µl). Concentrations of analyzed compounds in the starting solutions, as well as in the donor and acceptor solutions after incubation were determined using an HPLC method. The first PAMPA model was selected for the estimation of passive gastrointestinal absorption of the remaining seven compounds. Derivatives 13 and 14 showed the lowest permeability coefficients (logPe values were -4.53 and -4.4, respectively). On the other hand, the highest permeability was determined for derivative 7 (logPe = -3.94). The highest membrane retention was observed for compound 6 (34%) and compound 3 (24%). This could explain lower permeability of these compounds than expected based on their lipophilicity.