Water surface temperature (including sea surface temperature and lake water surface temperature) is a key parameter for studying global or regional climate change and numerical weather prediction (NWP), as well as being an essential controlling variable in the exchange of heat, moisture and gases between water surface and atmosphere. It has important significance for understanding the biophysical processes of water body. Satellite measurements of water surface temperature are well established with 30 years' collection of practical data and have incomparable advantages over traditional observations. Their limitations and challenges are also identified at the same time. There are more than 30 types of infrared/microwave radiometers which can be used for measuring SST/LWST, and their resolutions, advantages and disadvantages are summarized and compared in this paper. SST/LWST measurements depend on a combination of atmospheric properties and water surface radiances. Therefore, it is necessary to adjust and correct the atmospheric effect and water surface processes. The basic principles and the main types of algorithms for water surface temperature inversion using infrared and microwave data are illuminated and reviewed briefly. There are many uncertainties associated with SST/LWST measurements, and the magnitude of these uncertainties has put restrictions on the application or interpretation of SST/LWST measurements. A detailed analysis about these uncertainties in both infrared and microwave SST/LWST retrieval including undetected cloud, water vapor, aerosols, emissivity and skin effect is conducted. In order to determine the uncertainties in satellite-derived surface temperature, the validation of surface temperature retrieval is an indispensable step. Finally, a prospection about the trend of water surface temperature retrieval is proposed. Additionally, a strategy is advised for assimilating measurements from multi-sensor data in order to take the advantage of their complementary strengths.