The aim of this study was to test the ability of an infrared (IR) camera to assess temperature changes and distributions in teeth below restorations when quartz-tungsten-halogen (QTH) and light-emitting diode (LED) curing lights were used to photopolymerize the restorative material. Our hypothesis was that the higher power density and broader spectral distribution of the QTH source would cause greater increases in tooth temperature than the LED source, and that these differences would be best demonstrated with the IR camera. Cavities were prepared on human third molars and restored with a resin composite restorative material. The material was light-cured using three light-curing sources using several exposure times. The external (outside the tooth) and internal (inside the pulp chamber) temperature changes during polymerization of the composite material were recorded over 360 s with thermocouples and an IR camera. Using thermocouples the maximum increase in external temperature (+17.7 degrees C) was reported for the Swiss Master light after 20 s of curing time while the minimum temperature rise (+7.8 degrees C) was reported for the Freelight 2. Whereas a 2.6 degrees C increase in internal temperature was observed after curing 20 s with the Freelight 2, 7.1 degrees C was reported after 60 s of light exposure to Astralis 10. Infrared images showed similar trends in external-internal rises in temperature as the thermocouples, although temperatures measured by the IR were generally higher. These results indicate that the higher power density QTH sources caused greater increases in tooth temperatures than the LED source and that thermocouples may underestimate the heat applied to the tooth.