BACKGROUND: Air Quality indicators or indices (AQIs) are mainly used for communicating the air pollution levels and risk to the general population. However, very few epidemiological studies have used AQIs for characterizing exposure.OBJECTIVE: In the framework of the RESPOZE panel study we evaluated the association of daily ozone AQI levels with the daily occurrence of respiratory symptoms and Peak Expiratory Flow (PEF) and compared the effects with those estimated using measurements from fixed outdoor monitoring sites, in the city of Athens, Greece.MATERIALS AND METHODS: A panel of 97 children, aged 10-11years, was followed intensively for 35days (5weeks) during the academic year 2013-14. PEF and symptoms were recorded daily by each child. Two ozone AQIs classifying the air quality into 7 categories of increasing severity, were calculated; one characterizing the whole Athens area and one the local area around the child's residence and school. Measurements from fixed sites were also used. Mixed effects models for repeated measurements were applied, adjusting for several confounders.RESULTS: Increasing ozone levels were associated with increased incidence of symptoms, but the strongest and most statistically significant associations were found with the local air quality characterization with the AQI. Specifically, an increase in AQI-local by one category was associated with 34% (95% CI: 9%, 64%) increased odds of stuffy nose. When the AQI categories were "Bad" and "Severe", an increase in the incidence of cough was observed (OR 3.05 (95% CI: 1.29, 7.22) and 6.42 (95% CI: 1.47, 28.03) respectively). We did not observe a statistically significant association between AQI and PEF.CONCLUSION: Our results show that the use of an AQI based on local conditions may be advantageous over the use of only measurements when investigating the effects of air pollution on health outcomes for improving communication of risk to the public.