Several commercially available spirometers use unheated ceramic elements as flow sensors to determine flow and calculate volume of air. The usual method of correcting the resulting flow and volume values to body temperature pressure saturated (BTPS) is to apply a constant factor approximately equal to 30 percent of the full BTPS correction factor. To evaluate the usual BTPS correction factor technique, we tested several sensors with a mechanical pump using both room air and air heated to 37 degrees C and saturated with water vapor. The volume signals used to test the sensors were volume ramps (constant flow) and the first four American Thoracic Society (ATS) standard waveforms. The percent difference in FEV1 obtained using room vs heated-humidified air (proportional to the magnitude of the BTPS correction factor needed) ranged from 0.3 percent to 6.2 percent and varied with the number of maneuvers previously performed, the time interval between maneuvers, the volume of the current and previous maneuvers, and the starting temperature of the sensor. The temperature of the air leaving the sensor (exit temperature) showed a steady rise with each successive maneuver using heated air. When six subjects performed repeated tests over several days (each test consisting of at least three maneuvers), a maneuver order effect was observed similar to the results using the mechanical pump. These results suggest that a dynamic, rather than static, BTPS correction factor is needed for accurate estimations of forced expiratory volumes and to reduce erroneous variability between successive maneuvers. Use of exit air temperature provides a means of estimating a dynamic BTPS correction factor, and this technique may be sufficient to provide an FEV1 accuracy of less than +/- 3 percent for exit air temperatures from 5 degrees to 28 degrees C.