This investigation analyzed the thermal comfort (according to international standards) and the performance (throughRe asoning Test Batch (BPR-5)) of students in the classroom during three consecutive days (18, 19 and 20, July, 2012).
The tests were performed at three levels of dry bulb temperature, respectively, 24°C, 20°C and 30°C. Was used the
statistical distribution of Kruskal-Wallis (α=0,05) to compare the samples. It was concluded that (1) only evidence
relating to the mechanical reasoning tests (Rm) showed significant variation (p=0.03464 <0.05); (2) there was
significant v ariation only in performance for (Rm) (p=0.0089<0.05) and numerical reasoning (Rn) (p=0.0036<0.05)between the temperature ranges in which students obtained better performance in the session at 30°C; and (3) there wasno significant reduction in the performance of students regarding the variation of air temperature, considering the tests individually.
Keywords: thermal comfort; performance, students