Presented at the 2018 Ocean Salinity Science Team and Salinity Continuity Processing MeetingWe present comparisons between satellite sea surface salinity (SSS) products from the SMOS, Aquarius and SMAP missions and assess some of the possible reasons for the observed differences. We also compare satellite products to in situ observations. The comparisons focus on SSS large scale spatial patterns, temporal variability at regions of reference and statistical distribution. We also assess the dependence of SSS differences (between satellite products and between satellite and in situ) to sea surface temperature (SST), which has been an ongoing issue in the previous product versions. In order to assess the possible causes for the differences in satellite SSS, we reprocess Aquarius retrievals using the dielectric constant model and ancillary sea surface temperature product used for SMOS. Then we quantify their impact on the differences between SMOS and Aquarius SSS. We also compare the satellite SSS products to in situ salinity measurements to assess the performances of various retrieval parameterizations. We will discuss the improvements brought in the latest versions of the Aquarius (V5) and SMAP (V2.1 ~ V3) products, in particular regarding the large scale biases and their SST dependence. The latest versions of SMAP and Aquarius products show improved accuracy and significantly lower dependence of the error on SST. SMAP V2.1 differences with in situ SSS are reduced by 0.1 - 0.15 psu compared to V2 and the strong dipole low/high latitudes in the spatial distribution of the error is largely removed. We will also discuss the changes in biases in coastal regions.