This study uses statistical downscaling to estimate the impact of future climate change on air quality. We employ historical observations of surface ozone (O3) over the Chicago area, large-scale climate variables from the National Center for Environmental Protection (NCEP) reanalysis data, and climate projections from three GCMs (GFDL, PCM, and HadCM3), driven by two SRES emission scenarios (AlFI and B1 for GFDL and PCM; A2 and B1 for HadCM3). This approach calculates historic relationships between meteorology and O3, and considers how future meteorology would affect ground-level O3 if these relationships remain constant. Ozone mixing ratios over Chicago are found to be most sensitive to surface temperature, horizontal surface winds, surface relative humidity, incoming solar radiation, and cloud cover. Considering the change in O3 due to global climate change alone, summertime (June, July, and August) mean mixing ratios over Chicago are projected to increase by 6-17 ppb by the end of the century, depending on assumptions about global economic growth and choice of GCM. Changes are greater under higher climate emissions scenarios and more sensitive climate models (e. g. 24 ppb for GFDL AlFI as compared to 2 ppb for PCM B1). However, this approach does not take into account changes in O3-precursor emissions nor changes in local and lake-effect meteorology, which could combine with climate change to either enhance or diminish the projected change in local mixing ratios. Statistical downscaling is performed with the Statistical DownScaling Model (SDSM v. 4.1, a publicly available scientific analysis and decision-support tool.