Limited studies were focused on primary school buildings especially under subarctic climate. Thermal comfort of children was assumed to be similar as that of adults, which may cause inaccuracy. To fill data blank and enrich global database, a field study was performed from late fall 2016 to early spring 2017 covering whole heating period in north part of Sweden. Indoor CO₂ concentration was continuously monitored to evaluate indoor ventilation. Thermal comfort related parameters were continuously measured and predicted mean vote (PMV) was calculated. Subjective questionnaire surveys were performed every week except holidays. Subjective thermal sensation value (TSV) was always higher than objective PMV, which reflected thermal adaptation. The thermal adaptation became not obvious in middle and late winter because of long term exposure to heating environments. Heating system should be intensified gradually in early heating period, operated based on actual outdoor climate instead of experience in middle and late heating periods, extended under part load operation in early spring if necessary. The new 13─point TSV scale was pointed out by other researchers and tested inthis study, which can explore tiny TSV deviations from thermally neutral status and reflect more accurate thermal sensations.