Snow algae have been proposed to play a key role in climate change as they can reduce albedo (“bioalbedo”) and thus accelerate the melting of snow and ice fields. Although satellite-derived data has opened opportunities for larger scale observations, remote sensing of snow algae has been scarce and is methodologically challenging due to the presence of other light-absorbing impurities (LAIs). So far the studies on the role of LAIs in reducing albedo and increasing melting have been strongly focused on the Arctic ice sheets. The aims of the present study were to compare the relative impact of microalgae and other LAIs in reducing albedo in the snow of Fildes Peninsula, King George Island, Maritime Antarctica, using Spectral Mixture Analysis (SMA), which allows mapping sub-pixel fractions of multiple components in mixed pixels from satellite-derived data (Sentinel-2A). Also, the applicability of band ratios previously proposed for classifying snow algae (Red-Green band ratio) and impurities (Snow Darkening Index (SDI)) was tested and compared with SMA. Ground validation was made by characterizing the composition of snow algae (through chlorophyll a fluorescence) and by measurements of spectral absorption of solar radiation in red and green snow. SMA resulted a reliable method to classify snow algae and impurities with low amount of false positives (user accuracy 92–93%). However, omission error derived from dominant type (>50% abundance) confusion matrix was higher (producer accuracy for algae 63% and impurity 53%). In contrast, classification with band ratios resulted in large number of false positives (user accuracy for SDI 36%, for R/G 46%), and even higher omission error for R/G (producer accuracy 36%), whereas SDI had better producer accuracy (68%). SMA provided higher precision in separating dominant LAIs than the band ratios, which resulted in widely overlapping signals. Reduced albedo could be related with SMA-derived snow algae and impurity abundancies at albedo levels >45% for algae and >30% for impurities.