The problems of consistency of sorption, diffusion and pervaporation data have been analyzed from the viewpoint of the method of sorption measurements. Sorption of the water/ethanol mixture by sulphonated polyethylene of different capacities was studied isopiestically and by direct immersion in the solution. For a given method the sorption isotherms per sulphonic group were similar for all samples but, in contrast to what had been assumed before, both methods gave completely different results. It was found that direct sorption results were more consistent with pervaporation and diffusion data. It suggests that in pervaporation studies it is preferable to use this kind of data; use of isopiestic data may lead to serious errors. The practical consequences of this conclusion, such as the possible difference between pervaporation and vapor permeation, have been also discussed. (C) 2000 Elsevier Science B.V. All rights reserved.