The paper presents the algorithms for retrieving atmospheric temperature and moisture profiles and surface skin temperature from the high-spectral-resolution Atmospheric Infrared Sounder (AIRS) with a statistical technique based on principal component analysis. The synthetic regression coefficients for the statistical retrieval are obtained by using a fast radiative transfer model with atmospheric characteristics taken from a dataset of global radiosondes of atmospheric temperature and moisture profiles. Retrievals are evaluated by comparison with radiosonde observations and European Center of Medium-Range Weather Forecasts (ECMWF) analyses. AIRS retrievals of temperature and moisture are in general agreement with the distributions from ECMWF analysis fields and radiosonde observations, but AIRS depicts more detailed structure due to its high spectral resolution (hence, high vertical spatial resolution).