Denitrification has been studied using measurements of stratospheric HNO3 and N2O by the Airborne Submillimeter Radiometer (ASUR), operated on board the NASA DC-8 during SOLVE/THESEO 2000. Lidar measurements taken on board the same aircraft have been used to distinguish between temporary uptake of HNO3 in polar stratospheric clouds (PSCs) and denitrification events. To derive an NOy budget, ClNO3 data by balloonborne and ground-based Fourier transform infrared measurements and a model estimate of NOx + 2N2O5 have been considered. The HNO3 profiles of sporadic ASUR measurements without PSC coverage in January suggest that denitrification had started in the vortex core region by then. Vortexwide denitrification was found in mid-March 2000. Corrected for diabatic descent using the N2O measurements, a vortex-averaged NOy deficit between 1.2 ¿ 0.9 ppb at about 16 km altitude and 5.3 ¿ 2.7 ppb at about 20.5 km altitude was derived compared to December 1999, based on an observed decrease in HNO3 between 2.2 and 3.5 ppb during this time period. A shift in the NOy partitioning from HNO3 toward ClNO3 of about 0.4 to 0.7 ppb was observed in mid-March compared to December, indicating that chlorine deactivation was occurring. Comparisons with the SLIMCAT three-dimensional chemical transport model applying denitrification schemes based on ice and nitric acid trihydrate particles in equilibrium, respectively, reveal agreement within the error bars at higher altitudes (~19 km) but show discrepancies at lower altitudes (~16 km). It is suggested that more sophisticated denitrification schemes are needed to generally describe denitrification processes.