As part of the Eastern Pacific Investigation of Climate Processes program, from 2000 through 2003, the easternmost 95¿W Tropical Atmosphere Ocean (TAO) moorings were enhanced to provide time series of net surface heat flux, and the National Oceanic and Atmospheric Administration ship maintaining the 95¿W and 110¿W TAO lines was enhanced to monitor surface heat fluxes and atmospheric boundary layer structure. In this study we compare the ship-based and buoy-based radiative, bulk latent, and sensible heat fluxes, as well as the meteorological state variables used to compute the turbulent heat fluxes. The buoy net surface heat flux measurements appear to have an overall uncertainty near the target 10 W m-2, when careful attention is paid to the state variables. When hourly averaged data were unavailable, the telemetered daily averaged data were used in combination with an estimate of the mesoscale gustiness. In the eastern tropical Pacific a warm layer correction to account for stratification above 1-m depth was important only during the warm season (January--May) near the equator. These high-quality, cross-validated buoy heat flux time series are then used to assess the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR), NCEP/Department of Energy, and 40-year European Centre for Medium-Range Weather Forecasts reanalyses' surface heat fluxes. All reanalyses show that over warm water where deep convection is prominent, latent heat loss is too large and net solar radiation is too weak; conversely, in regions of stratocumulus over cool water, net solar radiation is too strong, and for NCEP/NCAR, latent heat loss is too weak.