Recent satellite data analysis and reanalysis projects have provided an unprecedented wealth of data sets relevant to surface energy budget in the Arctic Ocean. To assess how well we can reconstruct the variations of surface radiative fluxes used as boundary conditions to force sea ice models, the surface temperature, surface downwelling shortwave and longwave radiative fluxes of the (1) Cloud and Surface Parameter Retrieval (CASPR), (2) International Satellite Cloud Climatology Project (ISCCP-FD), (3) National Centers for Environmental Prediction Reanalysis II (NCEP-R2), and (4) European Centre for Medium-Range Weather Forecast 40-Year Reanalysis (ERA-40) were evaluated over the period November 1997 to September 1998. At the local scale the accuracy of these surface parameters was compared to the high-quality in situ measurements from the Surface Heat Budget of the Arctic Ocean (SHEBA) over the seasonal cycle and during particular storm events. The CASPR and NCEP-R2 are closer to the SHEBA for the surface temperature. The CASPR and ISCCP-FD are more accurate for the surface downward shortwave radiative fluxes. The ERA-40 agrees well with the SHEBA for the surface downward longwave radiative fluxes. At the basin scale the consistency of dominant spatial/temporal variability of these surface parameters across different data sets was examined. All data sets reproduce the patterns associated with the seasonal cycle, but they vary in their ability to capture patterns linked to synoptic variability.