We assessed the utility and accuracy of a temperature-sensitive pacifier in screening for fever in ill children under 2 years of age. Of 189 candidates for study, 83 (42%) did not use pacifiers, and of the 106 who did, 25 (24%) could not sustain a suck for five minutes of direct observation. Among those 81 children who could sustain five minutes of sucking, only two of 20 children with rectal temperatures above 100 degrees F (37.8 degrees C) were correctly identified as febrile. Furthermore, seven of eight children with temperatures of 102 degrees F (38.9 degrees C) or greater were incorrectly identified as being afebrile. There were no false-positive fever assessments in afebrile infants. The temperature-sensitive pacifier tested in this study does not accurately identify fevers in most infants who are shown to have fevers by rectal temperature determination. The use of this pacifier for screening fever in ill infants cannot be recommended.