A non-linear Newtonian iteration retrieval method has been successfully applied to TIROS-N Operational Ver-tical Sounder (TOVS) data process and simulated Atmospheric Infrared Sounder (AIRS) data inversion. Usually, the Newtonian iteration procedure starts with a first guess or initial state of atmospheric temperature and water vapor profiles, in this paper, a climate mean profile and a regression retrieval from radiances were used as the first guess in AIRS data inversion, the Root Mean Square Error (RMSE) of both temperature and water vapor mixing ratio retrievals shows that high vertical resolution and accurate atmospheric temperature and water vapor profiles are achieved using either a poor (a climate mean profile) or good (a regression retrieval) first guess. This illustrates the capability of high resolution IR sounder radiances in atmospheric temperature and moisture profiles retrieval, which is important when there is no prior information or sample data set of atmospheric temperature and moisture available.