Abstract: Land data assimilation is a new means for multi-source geospatial data integration. We first introduced the conception and composition of the land data assimilation system (LDAS). Data assimilation is the method to integrate observations with model simulation results. The observations can be both direct and indirect data, from different sources and with variant resolutions. The output from a LDAS is datasets of land surface variables with spatiotemporal and physical consistence. We then reviewed major LDASs under development, which include the Global and North American Land Data Assimilation Systems developed by NASA (National Aeronautics and Space Administration), European Land Data Assimilation System, and Land Data Assimilation System for West China. We think that the Land Data Information System derived from LDAS will play a very important role in the new generation GIS.