The Polar Ozone and Aerosol Measurement (POAM) III solar occultation instrument has been measuring water vapor at high latitudes since April 1998. Retrievals extend from 5 to 50 km, with 5--7% precision throughout the stratosphere and a vertical resolution of 1 (3) km in the lower (upper) stratosphere. Estimated systematic errors in the stratosphere are 10--15%. In this paper, we validate the POAM III version 4 stratospheric water vapor using correlative measurements from satellite, airborne, and balloon-borne platforms. The resulting comparisons show that POAM water vapor is high compared to correlative measurements in the middle to lower stratosphere. The satellite (Halogen Occultation Experiment (HALOE) and Stratospheric Aerosol and Gas Experiment (SAGE) II) comparisons also indicate a sunrise/sunset bias in the POAM data, with sunset (Southern Hemisphere) retrievals larger than sunrise (Northern Hemisphere) retrievals by 5--10%. In the Northern Hemisphere, POAM is approximately 5--10% high compared to all validation data sets between 12 and 35 km. At higher altitudes this difference decreases, such that POAM agrees with HALOE at 40 km and is lower by 10% at 50 km. In the Southern Hemisphere, POAM is 15--25% higher than HALOE below 35 km, with differences decreasing to 10% by 50 km. Similar differences are seen with SAGE II. Despite these systematic differences the POAM water vapor data are self-consistent and show no long-term trends in accuracy or precision. Statistical comparisons of the water vapor variability measured by POAM, HALOE, and SAGE II show very good agreement. The POAM data are therefore valid for scientific studies, and the science community is encouraged to use this unique data set.