The archived ozone profiles from the Halogen Occultation Experiment (HALOE) have already been corrected for the effects of the spectrally varying, interfering absorption due to aerosols composed of aqueous sulfuric acid, and agreement with correlative measurements in the stratosphere is generally excellent. However, comparisons of sets of coincident HALOE and ozonesonde profiles indicate occasional large differences at the lowest levels of the stratosphere. Most of those instances occur at altitudes just below a well-defined minimum in the 5.26 μm channel aerosol extinction profile, whose wavelength dependence is not represented by a sulfuric acid aerosol model. Further, when the aerosol extinction exceeds about 10-3 km-1, the aerosol correction to the ozone channel transmittances is both large and uncertain. After screening out the HALOE ozone profile segments whose corresponding aerosol/cirrus corrections are likely uncertain and after averaging the ozonesonde profiles into 2.5 km thick layers, we find that the HALOE ozone agrees, on average, to within 10% of their coincident ozonesonde measurements down to 100 hPa at tropical/subtropical latitudes and to 200 hPa at extratropical latitudes. A tightening of the coincidence criteria for the comparisons does not improve the mean differences for the sets nearly as much. Part of the variance of the paired differences was also accounted for when the ozonesonde profile values were integrated into those 2.5 km layers, prior to taking differences. This improvement is due mainly to the vertical averaging of the local, higher-resolution ozonesonde data, matching the lower resolution for HALOE ozone in the lower stratosphere. It is concluded that HALOE is providing accurate ozone profiles throughout the lower stratosphere, when its correction for interfering aerosols has been well characterized and when cirrus layers are not indicated. ¿ 1999 American Geophysical Union