We present an analysis of the temporal evolution of stratospheric constituents above the station of Dumont d'Urville in Antarctica (67°S, 140°E) from August 14 to September 20, 1992. Data sets include temperature profiles and H2O, ClO, O3, NO2, ClONO2, HNO3, N2O, and CH4 mixing ratios and aerosol extinction coefficients from 46 to 1 hPa measured by the Microwave Limb Sounder (MLS) and the Cryogenic Limb Array Etalon Spectrometer (CLAES) instruments aboard the Upper Atmosphere Research Satellite (UARS). At the station, aerosol extinction coefficients and O3 profiles are obtained by a lidar together with O3 profiles provided by sondes. Integrated O3 and NO2 column amounts are given by a Système d'Analyse par Observation Zénithale (SAOZ) spectrometer located at the station. Column O3 is also provided by the Total Ozone Mapping Spectrometer (TOMS) instrument aboard the NIMBUS 7 satellite, complemented with potential vorticity derived from the U.K. Meteorological Office assimilated data set and temperature fields provided by the European Centre for Medium-Range Weather Forecasts. Time evolution of these measurements is interpreted by comparison with results from the SLIMCAT three-dimensional chemical transport model. We show that the site is near the vortex edge on average and is alternately inside the vortex or just outside in the region referred to as the "collar" region. There are no observations of polar stratospheric clouds (PSCs) over the station above 46 hPa (∼18 km). In fact, PSCs mainly appear over the Palmer Peninsula area at 46 hPa. The rates of change of chemical species are evaluated at 46 hPa when the station is conservatively inside the vortex collar region. The ozone loss rate is 0.04 ppmv d-1 (∼1.3% d-1), which is consistent with other analyses of southern vortex ozone loss rates; chlorine monoxide tends to decrease by 0.03 ppbv d-1, while chlorine nitrate increases by 0.025 ppbv d-1. These negative ClO and positive ClONO2 trends are only observed in the collar region of the vortex where O3 amounts are far from near zero, and little denitrification is observed. Loss and production rates as measured by UARS are more pronounced than the ones deduced from the SLIMCAT model, probably because of the moderate model horizontal resolution (3.75° × 3.75°), which is not high enough to resolve the vortex crossings above Dumont d'Urville and which leads to a larger extent of denitrified air than indicated by the UARS data. The analysis also shows activated ClO inside the vortex at 46 hPa, a dehydrated vortex at 46 hPa, and rehydrated above, with no trace of denitrification in the lower stratosphere. Good agreement between coincident measurements of O3 profiles by UARS/MLS, lidar, and sondes is also observed. Finally, the agreement between UARS and SLIMCAT data sets is much better in the middle stratosphere (4.6 hPa) than in the lower stratosphere (46 hPa).