The Atmospheric Infrared Sounder (AIRS) is acting as a heritage and risk reduction instrument for the Cross-track Infrared Sounder (CrIS) to be flown aboard the NPP and NPOESS satellites. The hyperspectral nature of AIRS and CrIS provides high-quality soundings that, along with their asynoptic observation time over North America, make them attractive sources to fill the spatial and temporal data voids in upper air temperature and moisture measurements for use in data assimilation and numerical weather prediction. Observations from AIRS can be assimilated either as direct radiances or retrieved thermodynamic profiles, and the Short-Term Prediction Research and Transition (SPoRT) Center at NASA's Marshall Space Flight Center has used both data types to improve short-term (0-48h), regional forecasts. The purpose of this paper is to share SPoRT's experiences using AIRS radiances and retrieved profiles in regional data assimilation activities by showing that proper handling of issues—including cloud contamination and land emissivity characterization—are necessary to produce optimal analyses and forecasts.