Precipitation media consisting of a variety of scatterers such as raindrops, hail, graupel and snow cause random fluctuations and scattering of the incident wave. A random distribution of many discrete scatterers causes the precipitation structures to be modeled as random media. The physical properties of interfaces and dielectric properties of the scatterer and background medium are the major factors responsible for scattering behavior at microwave frequencies. Waves scattered by random media have the potential to remotely determine microphysical properties such as average size, shape, orientation and concentration of scatterers. Most clouds evolve with the initial growth of liquid cloud droplets. Their subsequent development can be via a warm- or cold-based process, or a mixed-phase process. Warm-based clouds are warmer than 0°C and free of ice. Precipitation formation is restricted to condensation and coalescence. Most tropical clouds are formed by a warm process. In a cold-based process, diffusional growth of ice crystals is the dominant growth mechanism. Cold processes are common in winter storms and anvil regions of thunderstorms. In the mixedphase process, the cloud development begins with the condensation and coalescence process, but subsequently involves an ice process as well. In general, the precipitation structure contains both liquid and ice with varying shape and size throughout. Multiparameter radar measurements based on dual-polarization and dual-frequency techniques (in addition to the conventional Doppler parameters reflectivity, velocity, and velocity spectrum width), can p1ai an important role in the remote sensing of precipitating clouds. These techniques are fairly well established and can yield significantly more microphysical information than the Doppler products. For example, that the differential reflectivity ZDR, specific attenuation rate at X-band (10 GHz) A, and specific differential propagation phase shift KDP are important measurables related to the microphysical evolution of the cloud. In particular, the vertical profiles of the above radar observables are closely related to precipitation content and water phase2' ' ' , while ZDR has been shown to be useful for differentiating between rain and ice . A coupled graupel melting and radar model together with aircraft 2D-PMS (Particle Measuring System) data has shown that the vertical profile of ZDR provides an indication of the onset and progression of melting ice into raindrops5'7. Hail detection using ZDR and dual-frequency ratio hail signal (DFR-HS) is well documented5. At long wavelengths, ZDR is an excellent estimator of the reflectivity- weighted mean axis ratio of the raindrops filling the radar resolution volume8'9. TheX-band specific attenuation is related approximately to the fourth moment of the raindrop size distribution . Also, S-band (3 GHz) KDp is nearly related to mass weighted axis ratio of the raindrop size distribution9. The above mentioned multiparameter radar observables show marked differences in the ice, melting ice, and rain regions because the shape, orientation and dielectric constant are distinctly different in each region. Active remote sensing techniques using radar are capable of mapping the vertical inhomogeneities in precipitation clouds, but is not effective for global precipitation retrieval, especially over ocean. Spaceborne radiometer measurements over both sea and land offer an obvious advantage. The uncertainties associated with top-of-atmosphere (TOA) brightness temperature (TB) measurements to infer cloud vertical structure can be reduced by multi-frequency radiometer techniques and/or simultaneous radar observations . In the microwave regime, rain and cloud drops are good absorbers (albedo 0.5), while ice is essentially nonabsorbing (albedo < 0.95). Passive observations of the precipitation media with a microwave radiometer can be broadly classified as emission-based, scattering-based, or a combination of both . Scattering-based observations are generally made at frequencies above 60 GHz, where thermal emission from the underlying rain layer is scattered away from radiometer field of view by the presence of high albedo ice or snow. Thus, the scattering-based methods are relatively insensitive to the background surface. However, since the ice is responsible for the cold TB, the rain rate estimate through a purely scattering-based method is indirect. Emission-based methods are generally made below 22 GHz, and rely upon increases in the thermal emission from rain and cloud over a radiometrically cool ocean in accordance with Kirchoff's Law. In the intermediate frequency range, both scattering and emission/aborption processes are equally important. However, it is difficult to separate ground effects such as reduced surface emissivity due to land surface wetting. Using the differences in polarization brightness temperature, Spencer et al. developed a method to estimate rainfall over the ocean. The upwelling brightness temperature is the net result of interactions within a radiometer's footprint including all of the particle types (rain, melting, ice, etc.) and the background medium (land or ocean). Hence, it is difficult to resolve the individual layers in a precipitation media by using just a single frequency and polarization state. Recently, there have been a number of radiative transfer models developed which incorporate multi-frequency, multi-polarization components'3'14"5"6'17"8 . These are vertically and angularly detailed plane-parallel radiative transfer models. In this review, we will discuss the multifrequency TB output of a plane-parallel radiative transfer model which uses both multiparameter radar and cloud model data as input. Multiparameter radar is capable of measuring both range resolved backscatter parameters (Z, ZDR, and linear depolarization ratio LDR), and range cumulative propagation parameters such as KDP and Ax. However, TB observations represent a vertically integrated effect through the cloud structure. Similarly, resolution in radiometer observations can be improved by using frequency diversity. Thus there is ample scope in improving the existing remote sensing techniques by combining both active and passive microwave methods. In this review, multiparameter radar observations and their interpretation for microphysical retrieval are discussed first, while radiative transfer model simulations using both radar measurements and cloud model results are presented in the following section. Both qualitative and quantitative aspects of the precipitation remote sensing is emphasized throughout the discussion.