Ozone loss rates in the lowermost part of the Arctic stratosphere (at potential temperature levels less than or equal to 375 K) in the period January and February 1993 are calculated using a chemistry-trajectory model and 30-day back trajectories. The results were compared with observations carried out during the first Stratosphere Troposphere Experiment by Aircraft Measurements (STREAM) in February 1993 in the Arctic lower stratosphere. Relatively low N2O and low O-3 concentrations were measured during STREAM, and O-3 loss rates of 8.0 (+/-3.6) ppbv d(-1) were calculated from O-3-N2O STREAM data in the vortex area. The average O-3 loss rate calculated by the model is 8.6 ppbv d(-1) (1.3% d(-1)), in agreement with observations. However, the calculated O-3 loss rate decreases to the lower value of the observed loss rates when taking into account N2O-Cl-y interrelations from different studies. Heterogeneous reactions on liquid sulfuric acid aerosols, in particular those involving the chlorine reservoir species ClONO2 and HCl, must be considered to explain the observed O-3 loss rates. Complete conversion of ClONO2 and HCl to active chlorine by heterogeneous reactions in the model occurs at temperatures less than or equal to 205 K under conditions with enhanced aerosol loading, and at temperatures less than or equal to 200 K with background aerosol levels. Since the trajectory temperatures were frequently below 205 K and occasionally below 200 K, the model results are (1) sensitive to the Cl-y level but relatively insensitive to the initial chlorine partitioning within Cl-y and (2) show significant O-3 loss at background aerosol levels, being only 1-2 ppbv d(-1) less compared to conditions with enhanced aerosol loading. We conclude that future O-3 loss in the Arctic lower stratosphere is quite sensitive to temperature changes, while it appears to be less sensitive to enhanced aerosol loading (e.g., by volcanic sulfate particles).