The osmometric behavior of Drosophila melanogaster embryos in permeabilized eggs was studied in a microscope diffusion chamber designed to impose a rapid change in osmotic environment at various temperatures. A numerical model of NaCl diffusion in the chamber predicted that radial variations in concentration arising from the presence of a thin film of solution at the top of the chamber were negligible. On the basis of transient electrical conductance measurements in the chamber, characteristic time constants for the change in concentration averaged over the chamber depth occupied by the eggs were 0.99, 0.77, and 0.60 min at 0, 10, and 20 °C, respectively. The chamber response was sufficiently rapid that the characteristic response of the embryo was not masked. Equilibrium volumetric behavior of the embryos indicated that they behaved as nearly ideal osmometers over the range of 0.256 to 2.000 osm, and followed the relation FV eq = 0.123 C −1 + 0.541, where FV eq is equilibrium fractional volume and C is osmolality. Nonlinear regression of volumetric data during osmotic contraction yielded an average L p of 0.722 μm/(min · atm) at 20 °C and an apparent activation energy Δ E of 8.11 kcal/mol. The coefficients of variation in the L p estimates among individual embryos were 38, 18, and 47% at 0, 10, and 20 °C, respectively. With the use of probability rules and a model for volumetric behavior during freezing, it was determined that the observed variability in L p (assuming Δ E is fixed) considerably broadens the transition range of cooling rates over which the predicted probability of intracellular ice formation goes from 0 to 1. However, experimental observations (21) show the actual transition range is even wider, indicating that there exist other important sources of variability which determine the event of ice formation in D. melanogaster embryos.