The High Altitude Airship (HAA) is an unmanned, lighter-than-air vehicle presently in an Advanced Concept Technology Demonstration (ACTD) phase of development. The ACTD prototype is being developed by the Lockheed-Martin Maritime Systems and Sensors division under the direction of the Missile Defense Agency. The HAA will operate above the jet stream at an altitude of about 65,000 ft (nearly 20 km) in a quasi-stationary position maintained by solar-powered propeller thrusters and serve as a long-duration platform for surveillance, communication, and weather observation activities. It has been long-recognized that the northern hemisphere mid-latitude environment at this altitude is characterized by a climatological minimum in wind speed. However, transient wind variations do occur, often with wind speeds that easily surpass the HAA's expected propulsion capacity of about 15 m/s. The overarching technical objective of this project is the development of an autonomous flight control system to provide a station keeping capability for the HAA. In view of the HAA's limited propulsion capacity, it will be essential for the HAA's flight control system to exploit knowledge of the current and predicted environmental wind field in order to remain on station. Motivated by the limited amount of observed stratospheric wind data, we examined a 20-yr record of 6-hourly NCEP/NCAR Reanalysis data over several CONUS locations in order to obtain a better understanding of the meteorological mechanisms responsible for high-wind events. A frequency analysis of the data revealed sharp peaks at the annual, 90-day, and diurnal time scales, with the annual mode generally being dominant. The annual mode peak typically occurs between November and March, and appears to be related to Stratospheric Sudden Warming (SSW). The data reveal that during these events, it is not uncommon for the wind velocity to increase from a background value of about 10 m/s to speeds of 40 m/s or more over a period of 1-2 weeks. In this paper we present a description of the NCEP/NCAR Reanalysis data and the methods used in our analysis. We also demonstrate—with examples from the 2006-07 northern hemisphere winter—the capability of the NCEP GFS to predict stratospheric wind speed in the 1-2 week time frame. This capability will be essential for providing the HAA with sufficient lead time for evasive maneuvers during operations.