Severe wildfires and heatwaves pose challenges to modelling streamflow in glacierized basins. Most hydrological models are based on empirical albedo relationships that do not account for how wildfire soot deposition and heatwaves can decrease snow/firn/ice albedos and may not well represent the duration of the refresh of albedo from summer snowfall. It can therefore be advantageous to update model states based on remotely sensed albedo. This research develops a framework for satellite albedo data assimilation into a hydrological model to diagnose the effects of assimilation on prediction of streamflow from a highly glacierized basin, the Athabasca Glacier Research Basin. The Cold Regions Hydrological Model, forced by station observations, was used to predict streamflow over 2017 to 2021. Two runs were performed with the same parameterization: one with albedo data assimilation (DA) and a control run without it (CTRL). Ninety-four albedo spatial estimates based on Sentinel-2 imagery were used for assimilation. Summer streamflow data from the WSC Sunwapta River gauge (07AA007) were utilized to assess whether albedo DA was beneficial for glacier streamflow prediction. DA substantially outperformed CTRL during the wildfire-impacted year of 2018 but did not improve model performance in other years. Nevertheless, overall streamflow prediction bias declined greatly with DA even though NSE accuracy was similar. These findings reveal that satellite albedo DA can be beneficial for modelling streamflow in glacierized basins during heavily wildfire-impacted years, but is not needed during normal years or seasons with extreme heatwaves if model forcings are reliable and the model represents processes faithfully.