A detailed reanalysis of the calibration procedures for the National Oceanic and Atmospheric Administration (NOAA) advanced very high resolution radiometer (AVHRR) based on thermal-vacuum test data was performed as part of the National Air and Space Administration/NOAA AVHRR Pathfinder Project. This effort, a followup to work by Brown et al. (1985), was motivated by the finding that the AVHRR instruments on several NOAA platforms have been routinely operated outside the range of thermal-vacuum test results, and thus one could not interpolate nonlinear corrections directly from earlier methods. These new calibration procedures permit calculation of nonlinear temperature corrections for any AVHRR operating temperature based on a second-order polynomial regression with a total calibration accuracy relative to an external calibration standard of less than two digital counts (¿0.2 ¿C). Such an improvement is quite important to the absolute accuracy of surface thermal fields, which are derived from these data utilizing various multichannel atmospheric water vapor correction schemes. We find systematic differences in the newly derived nonlinear correction results and those reported previously by Weinreb et al. (1990) and the original reference material in the various addenda to NOAA NESS Technical Memorandum 107 (Lauritson et al., 1979). Calibration results for various AVHRR radiometers show instrument-similar corrections for each band. Radiometers on NOAA platforms 8--12 demonstrate similar nonlinearities. ¿ American Geophysical Union 1993