Land surface temperature (LST), which reflects surface properties, is one of the key parameters in the physics of land surface processes from local through global scales. LST is very required in time and space for a wide variety of scientific studies and thermal infrared (TIR) remote sensing applications. Satellite TIR channels are very available for LST retrieval but only in clear skies. However, when the surface is obscured by clouds, the actual retrieved LST for the corresponding pixel is, or is contaminated by, the cloud top temperature. Lacking understanding of the complex relationships between clouds and LST, the estimation of LST for cloud-covered pixels poses a big problem and challenge for thermal remote sensing scientists. In the present paper, a review of algorithms and approaches related to LST retrieval for cloud-covered pixels from TIR data is presented, and the characteristics of each method are also discussed. Directions for future research to improve the accuracy of satellite-derived LST for cloud-covered pixels are then suggested.