Current methods for the cryopreservation of peripheral blood stem cell transplants (PBSCTs) have been developed empirically. Although the use of cryopreserved PBSCTs is successful and usually leads to rapid hematopoietic recovery, the freeze-thawing process is known to induce a significant amount of cell death. Furthermore, the infusion of DMSO, which is used to protect the cells against damage induced by freezing, can cause morbidity. Freezing methods may be improved by using a fundamental cryobiological approach, addressing the putative causes of cell injury during freezing and thawing. Different cell types may have different optimal cooling rates. Cooling rates above this optimum may cause ‘fast cooling damage', e.g. by lethal intracellular ice formation, while cooling rates below the optimum may lead to ‘slow cooling damage', These cryoinjuries relate to the osmotic changes during freezing and thawing, and the resulting fluxes of water and cryoprotectant across the cell membrane. Mathematic modeling of these osmotic events can be used to predict the optimal cooling rates for specific cell types. Woelders and Chaveiro* have developed a model that calculates the ‘compromise' cooling rate for every subzero temperature, resulting in non-linear freezing curves. In the present study, this model was applied to predict ‘optimal' freezing curves for PBSCTs with 10% and 5% DMSO, respectively, using values for the membrane permeability coefficients and related parameters as published earlier for cord blood HPCs. These predicted curves were tested empirically and compared to the presently used standard linear freezing curve. CD34+ selected and unselected PBSCs were cryopreserved using the standard or the new freezing curves. Post-thaw quality was evaluated by cell viability, CFU-GM formation and megakaryocyte outgrowth. With 10% DMSO, the use of the predicted optimal freezing curve compared to the currently used freezing curve resulted in increased post-thaw viability of CD34+ cells (mean±SEM; 78.4%±6.6% versus 72.0%±6.1% for unselected CD34+ cells and 92.0%±0.6% versus 83.9%±2.5% [p<0.01] for selected CD34+ cells), colony formation (40.7%±8.8% versus 30.1%±7.9% [p<0.01] for unselected CD34+ cells and 102.6%±8.0% versus 90.1%±11.9% for selected CD34+ cells), and megakaryocyte outgrowth (6.0±0.7 versus 3.9±0.6 [p<0.01] CD41+ MKs per seeded selected CD34+ cell). Also lowering the DMSO concentration to 5% resulted in improved post-thaw viability and functionality, comparable to the results obtained with 10% DMSO and the predicted optimal freezing curve. The results obtained with 5% DMSO were not improved by using the theoretically optimized freezing curve, suggesting that the cooling rate of the theoretically predicted curve for 5% DMSO may have been too high. Indeed preliminary experiments with a slightly slower non-linear freezing rate suggest that further improvement is possible. Our results indicate that the current cryopreservation method for PBSCT can be improved by applying theoretically optimized freezing curves. Infusion of less DMSO and more viable cells will likely improve the outcome of PBSCT. * Woelders and Chaveiro, Cryobiology 2004, 49; 258–271