A new technique using a differential scanning calorimeter (DSC) was developed to obtain dynamic and quantitative water transport data in cell suspensions during freezing. The model system investigated was a nonattached spherical lymphocyte (Epstein-Barr virus transformed, EBVT) human cell line. Data from the technique show that the initial heat release of a prenucleated sample containing osmotically active cells in media is greater than the final heat release of an identical sample of osmotically inactive or lysed cells in media. The total integrated magnitude of this difference, Deltaqdsc, was found to be proportional to the cytocrit and hence also to the supercooled water volume in the sample. Further, the normalized fractional integrated heat release difference as a function of temperature, Deltaq(T)dsc/Deltaqdsc, was shown to correlate with the amount of supercooled cellular water which had exosmosed from the cell as a function of subzero temperature at constant cooling rates of 5, 10, and 20 degrees C/min. Several important limitations of the technique are (1) that it requires a priori knowledge of geometric parameters such as the surface area, initial volume, and osmotically inactive cell volume and (2) that the technique alone cannot determine whether the heat released from supercooled cellular water is due to dehydration or intracellular ice formation. Cryomicroscopy was used to address these limitations. The initial cell volume and surface area were obtained directly whereas a Boyle-van't Hoff (BVH) plot was constructed to obtain the osmotically inactive cell volume Vb. Curve fitting the BVH data assuming linear osmometric behavior yielded Vb = 0.258V0; however, nonlinearity in the data suggests that the EBVT lymphocyte cells are not "ideal osmometers" at low subzero temperatures and created some uncertainty in the actual value of Vb. Cryomicroscopy further confirmed that dehydration was the predominant biophysical response of the cells over the range of cooling rates investigated. One notable exception occurred at a rate of 20 degrees C/min where evidence for intracellular ice formation due to a DSC measured heat release between -30 and -34 degrees C correlated with a higher end volume but no darkening of the cells during cryomicroscopy. For the cooling rate tested (5 degrees C/min) the cryomicroscopy data correlated statistically very well with the DSC water transport data. A model of water transport was fit to the DSC water transport data and the average (5, 10, and 20 degrees C/min) biophysical parameters for the EBVT lymphocytes were found to be Lpg = 0.10 micro m/min-atm, ELp = 15.5 kcal/mol. Finally, the decrease in heat release from osmotically active cells measured by the DSC during repetitive freezing and thawing was found to correlate strongly with the viability of the cells measured during identical freeze/thaw protocols with cryomicroscopy. This shows the additional ability of the technique to assess freeze/thaw injury. In summary, this DSC technique is a promising new approach for measuring water transport in cellular systems during freezing.