The invention discloses a drought index construction method based on passive microwave remote sensing. The method comprises the specific steps that an inter-diurnal mask and an inter-monthly mask are constructed, and influences of mixed pixels of edges of a glacier, a snow cover and a water body on AMSR-E data during temperature retrieval are eliminated; a land surface temperature retrieval model is constructed through 10.8GHHz and 89.0GHz vertical polarization channel Tb of AMSR-E and used for performing retrieval on land surface temperature, wherein Tb is brightness temperature data of passive microwave remote sensing; a microwave normalized difference vegetation index model is constructed based on a microwave polarization difference index (MPDI); a dry edge equation and a wet edge equation of an MTVDI model are constructed based on the established land surface temperature retrieval model and vegetation index model; and the MTVDI model is established to simulate and monitor drought throughout the country month by month. The method has the advantages that the defect that visible light/near-infrared remote sensing is liable to influences of meteorological conditions is effectively overcome, and a great improvement is made to a TVDI model.