The Ussing chamber diffusion system was used as a model to study the apparent permeability across porcine nasal mucosa of eight drugs and molecules with different physicochemical characteristics, namely insulin, lidocaine, nicotine, PEG 4000, propranolol, sumatriptan, melagatran and an amino diether. A weak correlation was found between the apparent permeability coefficients and the corresponding literature data on the fraction absorbed after nasal administration in humans. In the case of passively transported drugs, a closer correlation was found than for the substances where other mechanisms such as carrier-mediated transport or possible efflux were involved. Factors influencing the correlation between in vitro and in vivo data are discussed and the importance of electrophysiological control of the viability status of the excised mucosa is emphasised. Although caution has to be exercised in view of the limitations of the in vitro system, it seems to be a useful tool when evaluating different factors influencing permeability of nasal mucosa.