Microwave radiation with the inherent advantage of its ability to partially penetrate clouds is ideally suited for remote measurements of precipitation, especially over the oceanic regions. The retrieval problem is of great practical interest, as precipitation over the oceans has to be necessarily remotely sensed. More importantly, precipitation is also a crucial input in many weather and climate models. A downward looking space borne radiometer such as the TRMM's Microwave Imager (TMI), however, does not sense the surface rain fall directly. Rather, it measures the upwelling radiation coming from the top-of-atmosphere which depends on the total quantities of the parameters that can affect the radiation in the chosen frequency range. In addition to precipitation, the attenuation and augmentation of total cloud content and integrated precipitable water content by absorption and emission, respectively, affect the microwave brightness temperatures in various frequencies. In the present work, a systematic study has been conducted to investigate the effect of total cloud and precipitable water contents on surface rainrate retrievals from the TMI measured brightness temperatures (BT). While a Bayesian framework is used to assimilate radar reflectivities into hydrometeor structures, a neural network is used to correlate rain and cloud parameters with brightness temperatures. The correlation between the TMI brightness temperatures and the TRMM's Precipitation Radar (TRMM-PR) is used as the benchmark for comparison. A community developed software meso-scale Weather Research and Forecast (WRF) is used to simulate the cloud and precipitable water content, along with the surface rainfall rate for several rain events in the past. Four cases are considered: (a) TRMM PR's near surface rain rate is correlated with TMI brightness temperatures directly; (b) the total cloud and precipitable water contents along with surface rainfall simulated using WRF are correlated with TMI BTs; (c) similar to case (b) with near surface rainfall taken from TRMM PR measurements; (d) total cloud and precipitable water contents and the surface rain rate are corrected with PR vertical reflectivity profile in a Bayesian framework. The freely available QuickBeam software has been used for simulation of reflectivities at the TRMM PR frequency. The corrected data are then correlated with TMI BTs. Results show that surface rain fall retrievals can be radically improved by using TRMM PR vertical rain corrected total cloud and precipitable water content. © 2012 IEEE.