The overall objective of the study was to evaluate the indoor air quality at a middle school with an emphasis on characterizing baseline conditions. The focus of this article is on the relationship between occupancy and measured concentrations of carbon dioxide, and an evaluation of the use of carbon dioxide as a tracer for ventilation in the school. The school was characterized as having no health complaints, good maintenance schedules, no carpeting within the classrooms or hallways, and no significant remodeling, and its officials had agreed to allow the sampling to take place during school hours. Monitoring followed the guidelines recommended in the "Preliminary Draft: Conceptual Standardized EPA Protocol For Characterizing Indoor Air Quality in School Buildings." Four indoor locations including the cafeteria, a science classroom, an art classroom, and the lobby outside the main office, and one outdoor location were sampled for various environmental comfort and pollutant parameters for one week in February 1997. A consistent relationship between hourly occupancy and corresponding carbon dioxide concentrations was seen. Carbon dioxide concentrations in the cafeteria, art room, and lobby were within specified American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE) guidelines for comfort (< 1000 ppm). The science room had the highest concentrations (frequently exceeding 1000 ppm) due to high occupancy and non-functioning unit ventilators. Measured ventilation rates were within specified ASHRAE guidelines for the art room, cafeteria, and lobby. The science room, which relied on natural ventilation only, was not able to meet the ASHRAE guideline on one of the three days studied. The use of a completely mixed space, one compartment mass balance model with estimated CO2 generation rates and measured CO2 concentrations is shown to be a useful method for evaluating ventilation. Modeled effective ventilation, air changes per hour, and mixing factors reflected measured carbon dioxide concentrations and measured ventilation in each room. Mechanical ventilation afforded better mixing than natural ventilation. This study demonstrates the usefulness of collecting indoor CO2 and occupancy data when carrying out indoor air quality evaluations in schools.