As the important parameters for meteorological and climate research, temperature and water vapor profiles are of importance to the study of radiation transfer, vertical stability, convective available potential energy, precipitation and cloud formation and evolution. At present, the main remote sensing techniques for detecting temperature and water vapor profiles include laser radars, microwave radiometers and hyper-spectral resolution detectors. The instruments and algorithms for temperature and water and vapor profile retrieval by ground-based infrared hyper-spectral remote sensing are summarized. The basic principles of statistical retrieval algorithms and physical retrieval algorithms are presented emphatically. The technical features of different algorithms in retrieval accuracy, stability and computing time are analyzed in detail. Finally, the limitation and future trend of the current atmospheric temperature and water vapor profile retrieval by infrared hyper-spectral remote sensing are discussed.