Soil moisture is a key variable influencing a variety of land surface processes. Accurate estimation of spatio-temporally distributed soil moisture is one of the challenging issues in quantitative remote sensing. This paper briefly describes the major algorithms for retrieving soil moisture using optical, passive-microwave and active-microwave remote sensing, or their combinations. The optical algorithms have relatively low accuracy of retrieval, but good spatial and temporal resolutions. The typical algorithms include the Index-based approach and the soil thermal inertia-based approach. The passive-microwave algorithms have relative high accuracy but low spatial resolutions. It can be grouped into the retrieval approaches for soil moisture only and the approaches for relevant parameters in addition to soil moisture. The active-microwave algorithms have generally high accuracy with a high spatial resolution. The algorithms can be divided into three classes: empirical, physical and semi-empirical approaches. In addition, a number of algorithms have been proposed, which combines in particular optical, passivemicrowave, or active-microwave data. Because the algorithms often combine the advantages of the multi-sensors, they can achieve a high accuracy with a good spatial resolution. With the achievement of retrieval techniques, several global soil moisture data sets have been generated. The widely used data sets include the European Remote Sensing satellites/ Meteorological Operational satellite programme (ERS/MetOp) data sets, the Advanced Microwave Scanning Radiometer for EOS (AMSR-E) data sets, and the Soil Moisture and Ocean Salinity (SMOS) data sets. The ERS/MetOp data sets provides global soil moisture data with a spatial resolution of 25-km so far since July, 1991, retrieved from the TU-Wien approach using C-band microwave data. The AMSR-E data sets provides global soil moisture data with a spatial resolution of 25-km for the period from June, 2002 to September, 2011, retrieved from the Land Parameter Retrieval Model (LPRM) using C-band and X-band microwave data. The SMOS data sets provides global soil moisture data with a spatial resolution of 40-km so far since November, 2009, retrieved from the L-band Microwave Emission of the Biosphere model (LMEB) using L-band microwave data. To improve retrieval accuracy of soil moisture, the new satellite sensors are scheduled to be launched into space, for example, the Phased Array type L-band Synthetic Aperture Radar-2 (PALSAR2) in 2013 and the Soil Moisture Active Passive (SMAP) in 2014.