Although the advanced microwave sounding unit (AMSU) on board the NOAA 15 and NOAA 16 satellites is primarily designed for profiling atmospheric temperature and moisture, the products associated with clouds and precipitation are also derived using its window channel measurements with a quality similar to those derived from microwave imagers such as the Special Sensor Microwave Imager. However, the AMSU asymmetry in radiance along the scan was found to be obvious at its window channels and could severely degrade the quality of cloud and precipitation products if not properly corrected. Thus a postlaunch calibration scheme is developed for these channels, and the causes of the asymmetry are analyzed from the AMSU instrument model. A preliminary study shows that the asymmetry may be caused by either the AMSU polarization misalignment or the antenna pointing angle error. A generic radiative transfer model is developed for a single-layered cloud using a two-stream approximation and can be utilized for the retrievals of cloud liquid water (L) and total precipitable water (V), cloud ice water path (IWP), and particle effective diameter (De). At the AMSU lower frequencies the scattering from cloud liquid is neglected, and therefore the retrieval of L and V is linearly derived using 23.8 and 31.4 GHz. However, for ice clouds the radiative transfer model is simplified by neglecting the thermal emission, and therefore the retrieval of IWP and De is analytically derived using the AMSU millimeter wavelength channels at 89 and 150 GHz. These cloud algorithms are tested for the AMSU on board the NOAA 15 and NOAA 16 satellites, and the results are rather promising. It is also found that the AMSU-derived cloud ice water path is highly correlated with the surface rain rates and is now directly used to monitor surface precipitation throughout the world.