We have used an off-line chemical transport model (CTM) to investigate the effect of model resolution on the calculated ozone destruction in the Arctic lower stratosphere in winter and spring. The CTM was forced using the meteorology of the 1994--1995 winter for horizontal resolutions from 5.6¿¿5.6¿ to 1.4¿¿1.4¿. Using both a full chemistry scheme and a simple chemistry scheme based only on the ClO-ClO catalytic ozone destruction cycle, we find very little sensitivity of the polar ozone loss to model resolution. This lack of sensitivity is due to the effectively complete activation of chlorine at all model resolutions during this cold winter, which removes any potential effect of spatial averaging on the ClO distribution. Our model results are in strong contradiction with the recent calculations of Edouard et al. <1996>. The differences may be due to differences in tracer diffusion between the studies. Comparison of the filamentary structures produced in the ClOx fields from the simple and full chemistry schemes show that chemistry plays an important role in determining the lifetime of these features in the model. Narrow filaments, which are no longer apparent in the full chemistry integrations, persist for much longer in the simple chemistry case at the same resolution. ¿ 1998 American Geophysical Union