We tested a transcutaneous core temperature sensor using a method that relies on the principle of zero heat flow. We tested the hypothesis that transcutaneous and rectal temperatures would track within 0.3 degrees C of each other for >90% of the time. A thermistor was placed between the infant's abdomen or back and the incubator's or radiant warmer's mattress, or within the axilla, attached to the skin with a foam adhesive disk insulator. Thirty preterm infants were either placed on their abdomens or backs in a convective incubator or under a radiant warmer, and continuous transcutaneous and rectal temperatures were measured for 1 hour. There were no significant differences between abdominal and core temperatures or between axillary and core temperatures measured in double-walled convective incubators or in radiant warmers. The rectal-abdominal temperature difference was significantly less than the rectal-axillary difference (p < 0.02) in convective incubators, but not when the infant was placed prone under radiant warmers (p = 0.27). Transcutaneous thermometry is reliable for monitoring core body temperature as indicated by rectal temperature in stable preterm infants in a convective incubator.