A climatology of aerosol surface area inferred from satellite measurements is used as input in a two-dimensional model to study the long-term evolution of polar ozone depletion, especially the Antarctic ozone hole. It is found that volcanic aerosol inputs very likely modulate the severity of the ozone hole. In particular, the rapid deepening of the ozone hole in the early 1980s, as seen, for example, in the Halley Bay total ozone measurements, was probably caused by accelerated heterogeneous chemistry associated with an increase in aerosol surface area due to volcanic injection combined with the anthropogenic perturbation of stratospheric chlorine. This is further substantiated by the large Antarctic ozone decline observed and modeled after the eruption of Mount Pinatubo. A number of factors that influence the ozone hole are also investigated, including the effect of liquid versus frozen aerosol, the effects of denitrification and dehydration, the role of HOx in HCl and ClONO2 recovery, and the effect of chlorine partitioning at the start of winter. Denitrification tends to slightly increase modeled ozone loss, primarily between about 17 and 25 km late in the season, while dehydration tends to decrease the amount of ozone depletion. However, temperature and aerosol amount have the strongest control on the model ozone loss for a given chlorine loading. These findings suggest that future Arctic ozone depletion could be severe in unusually cold winters or years with large volcanic aerosol surface area. ¿ American Geophysical Union 1996