This study presents the evaluation of the performance and acceptability of thermal comfort by students in the classrooms of a university building with minisplit-type air-conditioning systems, in a tropical climate. To carry out the study, temperature and humidity measurements were recorded, both outside and inside the selected classrooms, while the students were asked to complete thermal surveys on site. The survey model is based on the template proposed by Fanger and it was applied to a total number of 584 students. In each classroom, the Predicted Mean Vote (PMV) and the Predicted Percentage Dissatisfied (PPD) were estimated according to Fanger’s methodology, as well as the Thermal Sensation Vote (TSV) and the Actual Percentage Dissatisfied (APD), which were obtained from the measurements and the surveys. The results of this study showed that the PMV values, although they may vary with the insulation of the clothing, do not affect the TSV. Furthermore, comparing PMV vs. TSV scores, a 2 °C to 3 °C difference in operating temperature was found, whereby the thermal sensitivity for TSV was colder, so it could be assumed that the PMV model overestimates the thermal sensitivity of students in low-temperature conditions. In addition, an acceptability by 90% with thermal preferences between 23 °C and 24 °C were also found. These results indicate that it is possible to increase the temperature set point in minisplit-type air-conditioning system from 4 °C to 7 °C with respect to the currently set temperatures, without affecting the acceptability of the thermal environment to the students in the building.