Surface solar fluxes in the NOAA/GCIP shortwave radiation budget product are compared to those derived from the ISCCP D1 data for January, April, July and October 1998 for the continental US. The scheme for estimating fluxes is essentially the same in these products, and it is the shortwave algorithm developed at the University of Maryland. Radiances observed by the GOES-8 satellite serve as input for both products, however the processing of satellite data, especially the cloud detection, is different. In addition, the spatial and temporal resolutions are different: fluxes in the NOAA/GCIP product are available on a 0.5-degree lat-lon grid at every hour, while those in the ISCCP-D1 SRB product are provided on a 2.5-degree lat-lon grid at every three hours. Surface fluxes from the NOAA/GCIP product are larger by about 5-17 Wm-2 when averaged over the US. The differences in the surface fluxes are analyzed in terms of the differences in the input data, primarily, top of atmosphere (TOA) albedo and cloud cover. Shortwave (SW) TOA albedo, which determines the aerosol and cloud reflection and absorption in the flux-retrieval process, is systematically smaller (by about 0.018-0.034 albedo units) than that in the ISCCP-D1 product, when averaged over the US. Area-average cloud cover in the GCIP product is also smaller (by about 0.1-0.33 fractions) than that in the ISCCP-D1 product. These inputs would lead to a much decreased surface solar flux in the GCIP product. However, cloud optical depth in the GCIP product is 2-3 times larger than in the ISCCP-D1 product, and thus partially offsets the increase due to the smaller cloud cover. The SW TOA albedos are also compared to those measured by the CERES instrument on TRMM satellite and to those from the ScARAB project.