In 1999/2000, as many as 51 Electrochemical Concentration Cell (ECC) ozonesondes were launched from December to March at the Canadian Arctic Eureka observatory (80 N, 86 W), one of the most northern stations in the Arctic, and the temporal evolution of the vertical ozone profiles was obtained in detail. During the winter, monthly average total ozone and temperature at 50 hPa, both observed with the ECC ozonesondes, were substantially lower than normal over Eureka since 1993. From December to March, Eureka was most of the time inside the polar vortex in the lower stratosphere (475 K isentropic surface level), except for early March. When Eureka was inside the polar vortex, very low temperatures were observed in the lower stratosphere, in accordance with the detection of Polar Stratospheric Clouds (PSCs) by Mie lidar between the 17 and 22 km level in January. However, PSCs were not observed over Eureka around the same altitude in February. Together with the observations of large HNO 3 -containing Corresponding author: Michio Hirota, Meteorological Research Institute, 1-1, Nagamine, Tsukuba, 305-0052, Japan. E-mail: mhirota@mrj-jma.go.jp ( 2003, Meteorological Society of Japan particles by the ER-2, and depleted HNO 3 by the UARS Microwave Limb Sounder, the lidar observations suggest the denitrification in the lower stratosphere in late winter inside the vortex. Under these conditions, the intravortex ozone mixing ratio on the 475 K isentrope decreased by about 2.2 ppmv from 3.1 ppmv on the 497 K isentrope (4 February), to 0.9 ppmv on the 475 K isentrope (29 March) in late winter to early spring, following the diabatic descent. In this period, the loss rate inside the vortex was found to be 0.041 G 0.008 ppmv/day (90% confidence interval). These findings and results are consistent with other observations (THESEO 2000/SOLVE campaign), and suggest that significant chemical ozone loss did occur in the lower stratosphere inside the polar vortex during the winter of 1999/2000.