This paper presents the estimation of snow depth over north-western Indian Himalaya using the 18.7H and 36.5H GHz channels of Advanced Microwave Scanning Radiometer-EOS (AMSR-E). The Microwave Emission Model of Layered Snowpacks (MEMLS) was used along with AMSR-E to understand the difference in the snow pack emitted and sensor received signals due to the prevailing topography. The study shows that the brightness temperature of AMSR-E and MEMLS are comparable at 18.7 GHz with some differences in their values at 36.5 GHz showing the sensitivity of this channel to the prevailing topography. Three years of AMSR-E data were used to modify the 1.59 algorithm to suit the terrain and snow conditions of the north-western Indian Himalayas. The retrieved snow depth is then compared with ground observations. Data from December to February 2003-2006 were used for the study of snow depths less than 1 m. The modified algorithm estimates the snow depth better than the old algorithm over the mountainous terrains of the north-western Himalayas.