We have used a 3D chemistry transport model to evaluate the transport of HF and CH<sub>4</sub> in the stratosphere during the Arctic winter of 1999/2000. Several model experiments were carried out with the use of a zoom algorithm to investigate the effect of different horizontal resolutions. Balloon-borne and satellite-borne observations of HF and CH<sub>4</sub> were used to test the model. In addition, air mass descent rates within the polar vortex were calculated and compared to observations.<br> <br> Outside the vortex the model results agree well with the observations, but inside the vortex the model underestimates the observed vertical gradient in HF and CH<sub>4</sub>, even when the highest available resolution (1°×1°) is applied. The calculated diabatic descent rates agree with observations above potential temperature levels of 450 K. These model results suggest that too strong mixing through the vortex edge could be a plausible cause for the model discrepancies, associated with the calculated mass fluxes, although other reasons are also discussed.<br> <br> Based on our model experiments we conclude that a global 6°×9° resolution is too coarse to represent the polar vortex, whereas the higher resolutions, 3°×2° and 1°×1°, yield similar results, even with a 6°×9° resolution in the tropical region.