This study assesses the impact of assimilating satellite-observed snow albedo on the Noah land surface model (LSM)-simulated fluxes and snow properties. A direct insertion technique is developed to assimilate snow albedo into Noah and is applied to three intensive study areas in North Park (Colorado) that are part of the 2002/03 Cold Land Processes Field Experiment (CLPX). The assimilated snow albedo products are 1) the standard Moderate Resolution Imaging Spectrometer (MODIS) product (MOD10A1) and 2) retrievals from MODIS observations with the recently developed Pattern-Based Semiempirical (PASS) approach. The performance of the Noah simulations, with and without assimilation, is evaluated using the in situ measurements of snow albedo, upward shortwave radiation, and snow depth. The results show that simulations with albedo assimilation agree better with the measurements. However, because of the limited impact of snow albedo updates after subsequent snowfall, the mean (or seasonal) error statistics decrease significantly for only two of the three CLPX sites. Though the simulated snow depth and duration for the snow season benefit from the assimilation, the greatest improvements are found in the simulated upward shortwave radiation, with root mean squared errors reduced by about 30%. As such, this study demonstrates that assimilation of satellite-observed snow albedo can improve LSM simulations, which may positively affect the representation of hydrological and surface energy budget processes in runoff and numerical weather prediction models.