Motivated by the importance of quantifying the regional surface heat balance over Arctic sea ice in studying climate processes, Lindsay and Rothrock (1994) developed a methodology for computing regional surface sensible heat fluxes using readily available advanced very high resolution (AVHRR) IR satellite imagery. Their technique is based upon the determination of the pixel-by-pixel sea ice surface temperature from which estimates of sensible heat fluxes are then made. We compare the sensible heat fluxes over the Beaufort Sea computed using their methodology with those measured by a gust probe system on the National Oceanic and Atmospheric Administration P-3 aircraft on April 18, 1992, during the Leads Experiment. We use an AVHRR image recorded during the P-3 flight at 2303 UTC. We show that individual lead heat fluxes can be large, 115 W m(-2) for 1-km average fluxes obtained from flight legs that included a 300-m-wide lead, but that regional values of sensible flux over 50-200 km of sea ice were small and positive, similar to 8 W m(-2). The sensible heat flux computed from the P-3 showed that a value of C-s = 1.1 x 10(-3), where C-s is the heat transfer coefficient relative to 10 m, is appropriate for the spring Beaufort Sea. We suggest that more realistic winds derived, for example, from the National Meteorological Center sea level pressure analyses be used instead of the constant value of 5 m s(-1) now employed by Lindsay and Rothrock. We also found that the maximum value of Delta T, the difference between air and surface temperature, used in the calculation of sensible heat flux using the Lindsay and Rothrock technique was underestimated by a factor of 1.9 when compared with direct measurements. Use of this Delta T correction factor, synoptic scale winds, and the calculated value of C-s gave a good comparison between the AVHRR approach and aircraft fluxes measured over the region. The effective regional momentum drag coefficient C-D relative to 10 m was 2.1 x 10(-3) typical of Arctic pack ice.