It has been well established that antifreeze peptides are responsible for the ability of many marine teleosts to survive in icy sea water at temperatures below the colligative freezing points of their blood. However, the in vivo site of antifreeze action has yet to be determined. One current hypothesis, the ‘ice-exclusion hypothesis’, suggests that antifreeze proteins act within epithelial membranes (skin, gills, gut) and block the entry of external ice crystals. This hypothesis was examined by measuring ice propagation temperatures across isolated skin samples from the winter flounder (Pseudopleuronectes americanus). The results obtained were consistent with the hypothesis, indicating that fish skin is an effective barrier to ice propagation and that the effectiveness of this barrier improves with the addition of antifreeze proteins. The results also demonstrated that the skin ice propagation temperatures were substantially lower than the lethal freezing temperatures of the fish. Therefore, some other epithelial tissue (possible gill) must be less effective than the skin in blocking ice crystal penetration into the fish