For the application of soil moisture and ocean salinity(SMOS)remotely sensed sea surface salinity(SSS)products,SMOS SSS global maps and error characteristics have been investigated based on quality control information.The results show that the errors of SMOS SSS products are distributed zonally,i.e.,relatively small in the tropical oceans,but much greater in the southern oceans in the Southern Hemisphere(negative bias)and along the southern,northern and some other oceanic margins(positive or negative bias).The physical elements responsible for these errors include wind,temperature,and coastal terrain and so on.Errors in the southern oceans are due to the bias in an SSS retrieval algorithm caused by the coexisting high wind speed and low temperature;errors along the oceanic margins are due to the bias in a brightness temperature(TB)reconstruction caused by the high contrast between L-band emissivities from ice or land and from ocean;in addition,some other systematic errors are due to the bias in TB observation caused by a radio frequency interference and a radiometer receivers drift,etc.The findings will contribute to the scientific correction and appropriate application of the SMOS SSS products. For the application of soil moisture and ocean salinity (SMOS) remotely sensed sea surface salinity (SSS) products, SMOS SSS global maps and error characteristics have been investigated based on quality control information.The results show that the errors of SMOS SSS products are distributed zonally, ie, relatively small in the tropical oceans, but much greater in the southern oceans in the southern Hemisphere (negative bias) and along the southern, northern and some other oceanic margins (positive or negative bias). physical elements responsible for these errors include wind, temperature, and coastal terrain and so on. Errors in the southern oceans are due to the bias in an SSS retrieval algorithm caused by the coexisting high wind speed and low temperature; errors along the oceanic margins are due to the bias in a brightness temperature (TB) reconstruction caused by the high contrast between L-band emissivities from ice or land and from ocean; in addition, some other systematic errors are due to th e bias in TB observation caused by a radio frequency interference and a radiometer receivers drift, etc. The findings will contribute to the scientific correction and appropriate application of the SMOS SSS products.