Air quality models used to predict the distribution and magnitude of photochemical smog pollutant concentrations require accurate input of meteorological and chemical components. The output from these models is strongly dependent on the accuracy of the horizontal and vertical structure of the wind field data used. In the Sydney region, surface air quality and meteorological data are available from several monitoring stations. Upper air data are available from the Bureau of Meteorology, obtained at Sydney Airport using a radiosonde system. This paper describes one selected high ozone day in the Sydney Basin and discusses the performance of meteorological models used to reproduce the atmospheric conditions important in predicting pollutant concentrations. The comparison was carried out between the observed data from Mascot and those predicted by the Bureau of Meteorology’s Limited Area Prediction System (LAPS) and The Air Pollution Model (TAPM) of CSIRO.