Cloudburst refers to the extreme form of precipitation, with a high amount of rainfall within a short span of time. It often leads to flash flood, landslide and damage of infrastructure due to heavy rainfall. The loss of lives is also not very uncommon. Hence a prediction mechanism of extreme precipitation is the need of the hour. The present paper presents a mechanism of prediction of cloudburst by detecting the formation of cumulonimbus type cloud, using the brightness temperature (TB) difference between 19 and 91 GHz TB values in horizontal polarization and descending passes of the SSMI satellite of NASA. This simple mechanism based on the Tb difference threshold value has been tested for several places in India, during 2013-2016. It is found that the prediction of cloudburst is possible with a lead time of 1-4 days using this algorithm.