Multi-window infrared measurements, together with visual channel observations, enable relatively high-resolution and accurate local, regional, and global retrievals of ocean surface temperatures to be repetitively and routinely obtained from operational environmental satellites. Drifting buoys appear to be the best means to date of validating the satellite estimates. Root mean square differences of about 0.6°C are found between satellite and drifter, whereas with ships-of-opportunity they are 1.8°C. Fixed buoy comparisons fall between these extremes. Abstract Multi-window infrared measurements, together with visual channel observations, enable relatively high-resolution and accurate local, regional, and global retrievals of ocean surface temperatures to be repetitively and routinely obtained from operational environmental satellites. Drifting buoys appear to be the best means to date of validating the satellite estimates. Root mean square differences of about 0.6°C are found between satellite and drifter, whereas with ships-of-opportunity they are 1.8°C. Fixed buoy comparisons fall between these extremes.