1] We describe the Polar Ozone and Aerosol Measurement (POAM) III NO 2 measurements and associated errors and compare the POAM III data to correlative measurements obtained from satellite-, balloon-, and ground-based instruments. POAM III NO 2 densities are retrieved from 20 to 45 km, with a vertical resolution of about 1.5–2.5 km at altitudes below 40 km and increasing to more than 7 km at an altitude of 45 km. Predicted random errors are on the order of 5% in this altitude range. Sunspots and high aerosol extinction can cause errors in the NO 2 retrievals but generally affect only about 10% of the data or less, depending on the altitude. The agreement between POAM III NO 2 data and correlative observations is excellent, demonstrating that the POAM III measurements are reasonable in terms of their magnitude, profile structure, and temporal variations. The largest number of comparisons was made with the Halogen Occultation Experiment (HALOE) on the Upper Atmosphere Research Satellite. On average, POAM and HALOE agree to within about 0.2 ppbv from 20 to 33 km or within about 6% at most of these altitudes, with no systematic bias. Differences increase to about 0.7 ppbv (17%, POAM higher than HALOE) by 40 km. This difference decreases to about 12% after accounting for a recently discovered error in the HALOE retrievals. Differences decrease above 40 km and are slightly negative (0.1–0.2 ppbv on average) at 45 km, the top edge of the valid POAM III NO 2 altitude range. We conclude that the POAM III NO 2 profiles from 20 to 45 km are appropriate for scientific analysis and for the validation of NO 2 measurements from other instruments.