Antarctic krill Euphausia superba (‘krill’) constitute a fundamental food source for Antarctic seabirds and mammals, and a globally important fisheries resource. The future resilience of krill to climate change depends critically on the winter survival of young krill. To survive periods of extremely low production by pelagic algae during winter, krill are assumed to rely partly on carbon produced by ice algae. The true dependency on ice algae-produced carbon, however, is so far unquantified. This confounds predictions on the future resilience of krill stocks to sea ice decline. Fatty acid (FA) analysis, bulk stable isotope analysis (BSIA) and compound-specific stable isotope analysis (CSIA) of diatom- and dinoflagellate-associated marker FAs were applied to quantify the dependency of overwintering larval, juvenile and adult krill on ice algae-produced carbon (αIce) during winter 2013 in the Weddell-Scotia Confluence Zone. Our results demonstrate that the majority of the carbon uptake of the overwintering larval and juvenile krill originated from ice algae (up to 88% of the carbon budget), and that the dependency on ice algal carbon decreased with ontogeny, reaching less than 56% of the carbon budget in adults. Spatio-temporal variability in the utilization of ice algal carbon was more pronounced in larvae and juvenile krill than in adults. Differences between αIce estimates derived from short- versus long-term FA-specific isotopic compositions suggested that ice algae-produced carbon gained importance as the winter progressed, and might become critical at the late winter-spring transition, before the phytoplankton bloom commences. Where the sea ice season shortens, reduced availability of ice algae might possibly not be compensated by surplus phytoplankton production during wintertime. Hence, sea ice decline could seriously endanger the winter survival of recruits, and subsequently overall biomass of krill.