The Ny-¿lesund Ozone Monitoring Intercomparison (NAOMI) took place at Ny-¿lesund, Spitsbergen (78.92 ¿N, 11.95 ¿E), from January 20 to February 10, 1998. This paper focuses on comparing stratospheric ozone profiles measured by the Alfred Wegener Institute differential absorption lidar (AWI DIAL), in routine Network for Detection of Stratospheric Change (NDSC) operation at Ny-¿lesund, the mobile Goddard Space Flight Center DIAL (GSFC DIAL), the University of Bremen microwave radiometer (μWave), and electrochemical concentration cell (ECC) ozonesondes, flown routinely by AWI. Below 30 km the two DIALs and the ECC sondes give virtually the same results, with instrumental precision (repeatability) better than ¿5% and no detectable bias. When their coarser altitude resolution is not accounted for, the μWave data show 15% low bias at 16 km and 15% high bias at 23 km. Considerably better agreement, better than ¿5% around 20 km and above 30 km, is found when the altitude resolution of the other data is degraded to match that of the μWave. During NAOMI the μWave data show high bias of up to 10% in a mixing ratio plateau around 25 km. Such bias has not been seen in routine intercomparisons between μWave and ECC sonde data at Ny-¿lesund. It is likely caused by an a priori profile 40% higher than the true profile during NAOMI. Above 30 km the μWave data show the best precision (repeatability), about ¿3 to ¿5%. Precision of the GSFC DIAL data decreases from better than ¿5% at 30 km to about ¿10% at 40 km, and the precision for the AWI DIAL data decreases from better than ¿5% at 30 km to ¿30% at 40 km. From 34 to 38 km the AWI profile is 12% lower than the GSFC profile. AWI DIAL measurements that are low at 35 km often end below 40 km or show high values at 40 or 45 km. This behavior seems related to the way in which the AWI processing algorithm changes altitude resolution for data with poor signal-to-noise ratio. ¿ 1999 American Geophysical Union