Background: We tested the hypothesis that classroom carbon dioxide (CO[subscript 2]) concentration is inversely related to child school attendance and educational attainment. Methods: Concentrations of CO[subscript 2] were measured over a 3-5?day period in 60 naturally ventilated classrooms of primary school children in Scotland. Concentrations of CO[subscript 2] were related to the class average annual attendance and proportions attaining a national standard for reading, writing, and numeracy, adjusted for socioeconomic status and class size. Results: The median (interquartile range, IQR) CO[subscript 2] concentration averaged over the school day was 1086?ppm (922, 1310). In the model, Time Weighted Average CO[subscript 2] concentrations were inversely associated with school attendance but not academic attainments. An increase of 100?ppm CO[subscript 2] was associated with a reduced annual attendance of 0.2% (0.04, 0.4) roughly equivalent to 1 half day of school per annum, assuming schools are open on 190?days per year. Indoor temperature and relative humidity were not related to attendance or academic attainment. Conclusions: Inadequate classroom ventilation, as evidenced by CO[subscript 2] concentration exceeding 1000?ppm, is not uncommon and may be associated with reduced school attendance. A relationship between inadequate classroom ventilation and adverse health outcomes in children may be present and this needs to be explored.