This paper investigates the variations in levels of nitrogen dioxide, NO2, monitored over the decade 2001-2010, in Newcastle-upon-Tyne (UK) city centre, to develop fundamental understanding of the periods of persistence of levels of NO2 greater than 40 μg m-3 (∼21 ppb) defined as air pollution event duration. The appropriateness of the hazard theory as a mechanism to understand failure rate of the duration of poor air pollution events was explored. The results revealed two types of air quality events. The longer duration air quality events (between 24 and 68 h) were associated with the ”extreme-weather” conditions and were responsible for a small number of extremely long air pollution duration events. These created bias in the results and therefore the analysis was restricted specifically to the 'normal-weather' related air pollution event durations, conforming to a geometric distribution. This novel approach shows promise as a mechanism to monitor and investigate year on year trends observed in air quality data.