We have measured stratospheric perhydroxyl (HO2) by using a sensitive mm-wave receiver to obtain spectroscopic line profiles of three rotational emission lines in the vicinity of 265.8 GHz. The observations were carried out over four days in September--October 1982 at Mauna Kea, Hawaii (19.5 N latitude) and yield good agreement with the column density and vertical distribution predicted above ~35 km by three representative two-dimensional photochemical models employing JPL 82-57 reaction rates and chemistry; predictions from the same models using WMO/NASA chemistry yield poor agreement with our observations. Contrasts between current theoretical predictions and previous observations by Anderson et al. are pointed out for HO2 in the 28 to 37-km range, along with the difficulty of joining the latter with our own measurements through a monotonically varying vertical profile for HO2. A possible explanation involving strong temporal and spatial variation in stratospheric water vapor content is suggested.