The albedo of land surfaces is an important parameter for surface energy balance and is dependent on the solar zenith angle and the three-dimensional structure of the canopies. The Moderate Resolution Imaging Spectroradiometer (MODIS) bidirectional reflectance distribution function (BRDF) algorithm is reformulated to derive a new two-parameter scheme for the computation of land surface albedo and its solar zenith angle (SZA) dependence for use in weather and climate models as well as in the remote sensing retrieval of surface solar fluxes. In this formulation the season- and pixel-dependent black-sky albedo at 60¿ SZA can be directly prescribed using the MODIS BRDF data while the two parameters are taken as a function of vegetation type only. Comparison of this formulation with those used in weather, climate, and data assimilation models (at NCAR, NCEP, and NASA) reveals the deficiencies in the treatment of SZA dependence and white-sky albedo in these models. Similarly, comparison of this formulation with those computed using the remotely sensed solar flux data from three groups (International Satellite Cloud Climatology Project (ISCCP)-FD, NASA's Clouds and the Earth's Radiant Energy System Tropical Rainfall Measuring Mission (CERES/TRMM), and University of Maryland) shows that surface albedos from these data sets are at times unrealistic, and suggestions are made regarding how to better treat the surface albedo and its SZA dependence in the retrieval of surface and atmospheric solar fluxes.