Radiometric performance of MODIS and VIIRS sensors is superior to that of the AVHRR, thanks to improved design and implementation of stringent pre-launch sensor characterization efforts and in-flight monitoring practices. Nevertheless, the imagery of the measured brightness temperatures (BT) and derived sea surface temperatures (SST) from multi-detector MODIS and VIIRS instruments is subject to striping artifacts. A robust adaptive destriping algorithm recently introduced by Bouali and Ignatov1 was optimized and operationally implemented at NOAA to remove striping artifacts in the VIIRS BT data. Destriped BTs are used as input into the NOAA Advanced Clear-Sky Processor for Oceans (ACSPO) SST system. The algorithm is also run with MODIS data onboard Terra/Aqua, in an experimental mode. We demonstrate improved image quality of VIIRS and MODIS BTs in bands centered at 3.7, 11 and 12 μm, and significant improvements in the derived SST imagery. The algorithm proves capable of removing the striping noise, while preserving the fine natural contrasts of the original satellite imagery. It was also tested to remove striping artifacts from the VIIRS and MODIS “optional SST” bands, centered at 4 and 8.5 μm. Destriping is critically important for several SST applications relying on accurate BT or SST gradient data, including ocean front detection and pattern recognition improvements to ACSPO cloud mask. We present the results of statistical characterization of striping artifacts in the VIIRS and MODIS thermal IR bands under various observational conditions. Our implementation of destriping is computationally efficient, adding only a fraction of time to the SST data processing flow. It is currently used at NOAA with VIIRS operations and reprocessing efforts.