Preliminary data on the use of infrared emission detection (IRED) tympanic thermometers suggest that operator technique may be important in IRED readings. No systematic investigation of specific technique and available IRED devices has previously been performed. In a prospective, blinded trial, 40 healthy adult subjects using six IRED thermometers with two techniques were examined in random sequence. Differences between IRED tympanic, oral, and rectal temperatures were compared using ANOVA. Significant differences were observed between all temperatures, the IRED devices, and the method of probe insertion. Differences between oral or rectal temperatures and IRED tympanic readings were reduced by an ear tug (as for routine otoscopy) for all but one device. An "ear tug" results in increased IRED readings that may improve accuracy of tympanic thermometers using IRED.