Among proteins that accumulate in plants in response to dehydrative forces or low temperature, dehydrins (late embryogenesis abundant [Lea] D11 family) have been the most commonly observed. Dehydrins are composed of several typical domains joined together in a few characteristic patterns, with numerous minor permutations. These domains include one or more putative amphipathic a-helix forming consensus regions, a phosphorylatable tract of Ser residues, and an N-terminal consensus sequence. Lesser conserved domains are also present at various positions, particularly between the putative a-helix forming domains, where they may occur as tandem repeats. This medley of permutations is mirrored by a wide size range of dehydrin polypeptides from less than 100 to nearly 600 amino acid residues. As of yet, the fundamental biochemical mode of action of dehydrins has not been demonstrated, but a number of immunolocalization and cell fractionation studies have established that dehydrins can be located in the nucleus or cytoplasm. Furthermore, it appears that these proteins associate with macromolecules ranging from nucleoprotein complexes in the nucleus to an endomembrane sheath in the cytoplasm. At present, all observations are consistent with a hypothesis that dehydrins are surfactants capable of inhibiting the coagulation of a range of macromolecules, thereby preserving structural integrity.