The effect of cooling rate and subsequent warming rate on survival of lactose-limited Escherichia coli was investigated. As previously reported, in the slow cooling rate range, a peak of survival was noted at 8 degrees C/min with survival decreasing as the cooling rate was increased or decreased from this value. Minimal survival was noted at 100 degrees C/min; increasing the cooling rate above 100 degrees C/min increased survival. At cooling rates greater than 200 degrees C/min, the survival became dependent on subsequent warming rates. Permeability damage, as measured by release of UV-absorbing material, potassium and beta-galactosidase, and increased accessibility of glucose-6-phosphate dehydrogenase to its substrates, was dependent on the cooling rate when cells were frozen in either water or saline. For cooling rates less than about 8 degrees C/min, there was minimal permeability damage to cells frozen in water. However, at rates greater than this value, damage and viability were related; the lower the viability the more the damage to the permeability barrier. The relationship was strengthened by the observations that protectants which increased survival reduced damage as well and that at ultrarapid cooling rates where survivals were dependent on warming rates, the extent, of damage was likewise dependent on the warming rate. Saline frozen cells were damaged by freezing and thawing more than comparable water-frozen cells over the whole cooling rate range. At cooling rates less than 8 degrees C/min, frozen in water, permeability damage of cells frozen in saline increased as the cooling rate decreased. As the cooling rate was increased from 8 degrees C/min, the damage increased as viability decreased. The relevance of these findings to the two-factor hypothesis of cell death is discussed.