The application of capillary flow porometry by gas-liquid displacement to the measurement of the pore size distribution in identical glass microfiber filter media can lead to surprisingly divergent results. The causes for these differences as well as the factors that influence the over-all reliability of data obtained by this widely used technique are investigated. Among the key factors studied were the volatility and viscosity of four common wetting liquids, the scan rate (i.e. the holding time between increments of differential pressure Delta p or volumetric flowrate V), and the scan sequence (i.e. dry before wet, or wet before dry scan). Most measurements were made with a porometer designed in house, in order to have complete control over all aspects of operation. Data obtained with commercial porometers are also reported. For best comparability, all measurements were made with the same batch of standard glass microfiber media. The largest error source by far was the volatility of fluorinated compounds commonly used as wetting liquids. While the vapor pressures of such compounds may be relatively low, their use in combination with a flow of air through the porous matrix can have an enormous effect on the evaporation rate during a scan. Neglecting this effect (which obviously depends on the scan rate) may ultimately result in an error of almost arbitrary magnitude in the pore size distribution. Silicone oil on the other hand has a negligible volatility and provides reliable results for a wide range of operating conditions. The liquid viscosity in the tested range of 5-100 mm(2)/s played a comparatively insignificant role. These and other factors of uncertainty are discussed on the basis of experimental data.