We have carried out a detailed analysis of several long pathlength transmission measurements in the 8–12-µm atmospheric window in order to determine the extinction coefficient due to the water vapor continuum. Our results indicate that three modifications to the current LOWTRAN atmospheric transmission model are required. The first two corrections involve an improved fit to the pure water vapor continuum absorption together with the elimination of the atmospheric broadened continuum term. Finally, and most critically, a strong measured temperature dependence must be included in the water vapor continuum absorption coefficient. For pathlengths ranging from 10 km to 50 km, failure to incorporate these corrections can lead to errors in the computed transmission ranging from factors of 2 to more than 10,000.