Aerosol surface area distributions inferred from satelliteborne 1-μm extinction measurements are used as input to a two-dimensional model to study the effects of heterogeneous chemistry upon anthropogenic ozone depletion at northern midlatitudes. It is shown that short-term (interannual) and longer-term (decadal) changes in aerosols very likely played a substantial role along with trends in anthropogenic chlorine and bromine in both triggering the ozone losses observed at northern midlatitudes in the early 1980s and increasing the averaged long-term ozone depletions of the past decade or so. The use of observed aerosol distributions enhances the calculated ozone depletion due to halogen chemistry below about 25 km over much of the past decade, including many periods not generally thought to be affected by volcanic activity. Direct observations (especially the relationships of NOx/NOy and ClO/Cly ratios to aerosol content) confirm the key aspects of the model chemistry that is responsible for this behavior and demonstrate that aerosol changes alone are not a mechanism for ozone losses in the absence of anthropogenic halogen inputs to the stratosphere. It is also suggested that aerosol-induced ozone changes could be confused with 11-year solar cycle effects in some statistical analyses, resulting in an overestimate of the trends ascribed to solar activity. While the timing of the observed ozone changes over about the past 15 years is in remarkable agreement with the model predictions that explicitly include observed aerosol changes, their magnitude is about 50% larger than calculated. Possible chemical and dynamical causes of this discrepancy are explored. On the basis of this work, it is shown that the timing and magnitude of future ozone losses at midlatitudes in the northern hemisphere are likely to be strongly dependent upon volcanic aerosol variations as well as on future chlorine and bromine loading. ¿ American Geophysical Union 1996