The aim of this study was to evaluate the increase in temperature induced by various light sources during in-office bleaching treatment, under simulated blood microcirculation in pulp conditions. Ten freshly extracted human maxillary central incisors were used for the study. The roots of the teeth were removed from approximately 2 mm below the cementoenamel junction and fixed on an apparatus for the simulation of blood microcirculation in pulp. A J-type thermocouple wire was inserted into the pulp chamber through an artificial access at the lingual surfaces of the teeth, and another thermocouple wire was fixed on the labial surface of the teeth meanwhile. An in-office bleaching agent, intense red in color and with 30 % water content, was applied to the labial surfaces of the teeth, and repeating measurements were made for each tooth using three different light sources: Er:YAG laser (40 mJ, 10 Hz, 20 s), 810-nm diode laser (4 W, 20 s, CW), and high-intensity light-emitting diodes (LED) (1,100 mW/cm(2), 20 s) as the control. Temperature increase in the pulp chamber and within the bleaching gel during light application were recorded and statistically evaluated. The highest pulp temperature increases were recorded for the diode laser group (2.61 A degrees C), followed by the Er:YAG laser (1.86 A degrees C) and LED (1.02 A degrees C) groups (p < 0.05; analysis of variance (ANOVA), Tukey's honestly significant difference (HSD)). Contradictorily, the lowest gel temperature increases were recorded for diode laser (6.21 A degrees C) and followed by LED (12.38 A degrees C) and Er:YAG (20.11 A degrees C) groups (p < 0.05; ANOVA, Tukey's HSD). Despite the significant differences among the groups, the temperature increases recorded for all groups were below the critical value of 5.6 A degrees C that can cause irreversible harmful changes in pulp tissue. It can be concluded that, with regard to temperature increase, all the light sources evaluated in this study can be used safely for in-office bleaching treatment within the described parameters.