As part of the Pathfinder program developed jointly by National Aeronautics and Space Administration (NASA) and National Oceanic and Atmospheric Administration (NOAA) a large database ofin situ sea surface temperature (SST) measurements coincident with satellite data is now available to the user community. The Pathfinder Matchup Database (PMDB) is a multi-year, multi-satellite collection of coincident measurements from the Advanced Very High Resolution Radiometer (AVHRR) and broadly distributed buoy data (matchups). This database allows the user community to test and validate new SST algorithms to improve the present accuracy of surface temperature measurements from satellites. In this paper we investigate the performance of a global Pathfinder algorithm to specific regional conditions. It is shown that for zenith angles less than 45°, the best-expected statistical discrepancy between satellite and buoy data is about ∼ 0.5 K. In general, the bias of the residuals (satellite — buoy) is negative in most regions, except in the North Atlantic and adjacent seas, where the residuals are always positive. A seasonal signal in SST residuals is observed in all regions and is strongest in the Indian Ocean. The channel-difference term used as a proxy for atmospheric water vapor correction is observed to be unresponsive for columnar water vapor values greater than 45 mm and high zenith angles. This unresponsiveness of the channels leads to underestimation of sea surface temperature from satellites in these conditions.