This paper describes a new steady permeation method for measuring water transport properties of fuel cell membranes. It is distinguished from most previous experiments of this type by employing a separate procedure for quantifying the extraneous sources of water transport resistance in the measurement cell that do not originate in the membrane (the so-called device resistance). This result is then used to determine the membrane water permeability in cells built with Nafion membranes in three thicknesses (25, 50, and 125 microns) operating at three different temperatures (70, 90 and 110 degrees C) over a range of relative humidities from 30-90%. The permeability data lie along three parallel straight lines (one for each temperature) on semi-log plots vs. average membrane relative humidity. There is no discernible dependence of permeability on membrane thickness, suggesting that any localized transport resistance that may exist at the membrane surface was too small to be detected by this method. The permeability data can be fit to a simple 3-constant empirical expression for the membrane permeability as a function of temperature and relative humidity. The largest difference between the present and previous measurements occurs at higher relative humidities, where the device resistance is large compared with the membrane resistance. (C) 2012 The Electrochemical Society. [DOI: 10.1149/2.016209jes] All rights reserved.