A three-dimensional off-line chemistry transport model, driven by European Centre for Medium-Range Weather Forecasts winds and temperatures, is used to interpret measurements of NO and NO2 taken from the DC-8 during the second Airborne Arctic Stratospheric Expedition. The model was run in three configurations: gas phase chemistry alone, inclusion of the N2O5 aerosol reaction, and inclusion of both N2O5 and ClONO2 aerosol reactions. The run including the N2O5 aerosol reaction alone usually agreed best with measured NOx/NOy ratios in midlatitude air masses. The NOx/NOy ratios of the run with both aerosol reactions were always too low, while the gas phase ratios were usually too high, especially during March. All three simulations generated extremely low NO2/NOy ratios in air parcels that had spent several days or more in the polar light. Measured NO2/NOy ratios in these types of air masses were sometimes equally low but could also be considerably higher. Observed NO/NO2 ratios differed strongly from known theory. ¿ American Geophysical Union 1994