The purpose of this study was to analyze pulpal temperature increase generated by prosthodontic margin repositioning and finishing with ultrasonic and rotating instruments. The temperature changes recorded were also correlated with the residual dentin thickness. A sample of 32 human extracted molars was selected. The teeth were endodontically treated and prepared with prosthetic round chamfer preparation. Then, they were inserted in plaster cubes up to the cement-enamel junction, leaving the apical portion pervious for inserting the thermocouple probe. The conventional technique, which involves the use of a high-speed contra-angle handpiece, was compared with an ultrasonic method (Crown Prep, Mectron, Carasco, Italy). For margin repositioning and finishing, two walls were randomly selected for each tooth: one was included in the test group and cut with the piezoelectric instrument (Multipiezo Touch TipHolder DB2, Mectron, Carasco, Genoa, Italy), the other one was inserted in the control group and cut with the high speed contra-angle handpiece (Kavo, Biberach, Germany). To standardize the operator-dependent parameters, it was used a mechanical arm controlled by a computer. These parameters were the pressure exerted on the dental wall, the cutting length and the time required for margin repositioning and finishing. For both test and control group, test phase consisted in a first stage of margin repositioning using an ultrasound tip or a diamond bur with a greater granulometry (120 µm for the ultrasound tip and 125 µm for the diamond bur), followed by a second finishing step conducted by an ultrasound tip or a diamond bur with smaller granulometry (60 µm for the ultrasound tip and 30 µm for the diamond bur). During these stages the intrapulpal temperature was recorded thanks to a thermocouple. Before and after these steps, the thickness of the remaining dental walls was measured with a caliber. The average pulpal temperature increase was 5.03±0.98 °C for the ultrasonic preparation (test group) and 3.55±0.95 °C for the conventional technique (control group). The difference was statistically significant (P value <0.001). However, neither of the instruments reached the critical level of 5.5 °C reported in the literature. The mean initial dentin thicknesses was 1.82±0.47 mm for the control group and 1.59±0.54 mm for the test group but the analysis of the residual dentin thicknesses revealed a greater reduction of the walls worked up with high speed contra-angle handpiece (mean 0.9±0.5 mm), which was therefore more aggressive than the ultrasonic instrument (mean 1.1±0.5 mm). A very weak negative correlation was present between the thickness of the wall at baseline and the increment of temperature. Within limitations of this study, temperature increasing of ultrasonic instruments shows a statistical difference related to rotary ones. But, as literature shows, the ultrasonic advantages are margin precision, preservation of soft tissues and reduction of operating times. Furthermore, in relation with results of this study, they could be considered safe for pulp vitality because the increase in pulpal temperature is similar to traditional instruments and it does not exceed the critical level of 5.5 °C.