Just as the term “climatology” evokes the concept of a characteristic mix of local meteorological variables that define a given site, the term “air quality climatology” is meant to evoke the concept of a characteristic mix of meteorological and air quality variables that define conditions at a site. The potential combination of variables for determining local air quality climatology is huge, but perhaps of greatest interest to air quality modeling is the mixture of wind direction and pollutant concentration. Air quality model applications typically include a performance evaluation that compares simulated pollutant concentrations against observations using a variety of statistical measures. Such evaluations produce quantitative measures of model performance for individual parameters such as the maximum daily 8-hour average ozone mixing ratio or the average daily PM2.5 concentration. Although this knowledge is important for understanding model skill in replicating the observed frequency distributions of pollutant values, it does not provide a complete assessment of model skill regarding its representation of local air quality climatology. A model that cannot accurately reproduce space-time (S-T) relationships between sources and receptors is all but useless for air quality management. One suggested method for testing the alignment between modeled and observed S-T pollutant behavior is through the use of two or three dimensional time-direction plots of concentrations. This is done by plotting the mean observed or modeled concentration field against time of day along one axis and wind direction along the other axis. Two dimensional concentration contours, that can be extended to three dimensional plots with the appropriate software, provide insight to the relationship between a pollutant and relevant space (with direction serving as a surrogate) and time variables. Such plots can be prepared using paired wind direction and concentration observations or paired direction and concentration model outputs. Any direct comparison of observed and modeled concentrations for a shared averaging period—as is typically done for the statistical comparisons that are now common in air quality modeling—ignores the spatial information inherent in wind direction. Examples are given of comparisons for ozone and aerosol concentration fields produced by the Models-3/CMAQ modeling system for sites in 2003 where continuous (hourly) air quality and wind observations allow detailed S-T analysis. Results indicate that modeled ozone performance can actually be better when using a direction-aligned comparison (i.e., the S-T plotting previously described), than if only a regression is done of paired hourly model outputs versus observations. However, modeled aerosol performance appears to worsen when direction-alignment is applied and, in general, is not nearly as good as ozone performance.