The oxidizing capacity of the atmosphere, defined as the global mean tropospheric abundance of the hydroxyl radical (OH center dot), strongly influences air pollution by controlling the lifetimes of gaseous pollutants and the production of particulate matter. Predicting future changes in OH center dot due to anthropogenic emissions and climate change is of interest to air quality managers, but it is difficult because of multiple competing effects. Models of atmospheric chemistry suggest that these competing effects buffer significant change in OH center dot in the past and in the near future. However, proxy-based observations for past changes in OH center dot and other oxidants over the preindustrial-industrial and glacial-interglacial time scales suggest much larger changes than models estimate. Model sensitivity studies show that variability in past and future OH center dot is highly sensitive to relative emissions of reactive nitrogen and carbon, water vapor, lightning, and stratospheric ozone, implying that one or more of these variables is highly sensitive to climate.