Fanger's PMV index has been widely used in the last ten years even if it has not yet been sufficiently tested in children. Furthermore recent studies have expressed doubts on the utilization of the index in real situations. We therefore studied the applicability of the PMV index via a survey that involved school children from 8 to 11 years old. The study was conducted comparing the individual objective quantification of thermal comfort (values of PMV index) and the feeling of subjective thermal comfort of people involved (by answering a standardized questionnaire). The PMV index values within the interval +/- 0.5 were grouped in 2 degrees C operative temperature classes; the percentages of subjects who expressed thermal comfort were calculated within these groups. Considering Fanger's assumption (as foreseen by ISO 7730), the percentage of subjects in a condition of thermal discomfort could not be higher than 10% in each group. On the contrary a percentage of dissatisfied persons was obtained that was significantly higher (p < 0.01) than 10% over all the temperature values considered. The number of discrepancies, analyzed even with respect to operative temperature values, was so high that it could not be attributed to the thermal variation that can be measured in the classrooms. Variations due to differences in the thermal resistance of clothing can be excluded since this was taken into consideration individually during the elaboration of the index.