Dependence of sensitivity on pressure in He, N2, and H2has been investigated for a group of 16 hot‐cathode ionization gauges, representing both extractor and Bayard–Alpert types, for total pressure as low as 5×10−8Pa. Absolute sensitivities were determined using a primary high vacuum standard. An independent method, that of measuring the sensitivity ratio of one gauge to another, was also employed. Within a scatter of about ±3%, ±4%, and ±4% respectively, the N2, He, and H2sensitivity ratio data showed no clearly discernible pressure dependence down to total pressures as small as 5×10−8Pa. The absolute sensitivity measurements in He and N2over the range 10−7to 10−3Pa in total pressure, also were constant within a scatter of about ±4% and ±3%, respectively. As a consequence of drift in the background component of the total pressure, all the absolute sensitivity measurements at total pressures below 10−7Pa exhibited an apparent pressure dependence not evident in the sensitivity ratio results. In the case of H2, drift in the pressure persisted at total pressures orders of magnitude larger than the background pressure, and for all the gauges led to an apparent difference of approximately 10% between the H2sensitivities at 10−8and 10−5Pa. Results of further investigation suggest that the apparent pressure dependence in the sensitivities is an artifact produced by the well‐known phenomenon of thermal dissociation of H2at hot filaments and associated processes of H2pumping and production of other species such as CO and C2H4.