The validity of oral temperature measurements depends on use of an appropriate site, a dependable thermometer, and a proper technique. Fifty febrile patients and 50 afebrile volunteers were studied to determine differences in oral temperature measured in three sublingual sites with three brands of electronic thermometers and with mercury thermometers and to learn whether two methods of probe insertion (slow-slide and direct-placement) with electronic thermometers would result in differences in either the temperature reading obtained or the instrument response time. When measured with the rapidly responding electronic thermometers, temperatures in the right and left posterior sublingual pockets were significantly higher (p<.01) than in the area under the front of the tongue. The difference was approximately .3° F in the afebrile subjects and approximately .4° F in those with fever. With mercury thermometers, however, the temperature differences were small and not statistically significant. There was no significant difference between temperatures measured in the two posterior pockets using either electronic or mercury thermometers. There was some individual variation in the values registered by electronic thermometers at each site, although the differences between the three sites were consistent with all three instruments. Actual temperature values differed between electronic and mercury thermometers. When electronic thermometer probes were inserted into a posterior pocket using a slow-slide technique, temperature readings were higher by approximately .1° F and were obtained a few seconds faster than when the probe was put directly in place without a prewarming procedure. These differences were statistically significant (p<-01) but of limited clinical importance. Findings emphasized the importance of using the posterior sublingual pocket as the site for oral temperature measurement and the value of using the same instrument to obtain repeated comparable measurements.