Studies on cell cryopreservation have been limited by the complexity of the freezing process and challenges on controlling ice formation, managing cooling rates, and optimizing cryoprotectant concentrations. This study objective is to evaluate the impact of bottom-up and conventional radial freezing on the viability of mammalian cells, using mouse hybridoma cells and human umbilical cord blood (hUCB) derived mononuclear cells (MNCs) as cell models. The study combines experimental assays, including cell viability assays and flow cytometry characterization, with Computational Fluid Dynamic (CFD) simulations. A bottom-up freezing geometry sustained high cell viability, even at dimethyl sulfoxide (DMSO) concentrations below 5% v/v, while using conventional radial freezing led to lower cell viability below such DMSO concentrations threshold. Note that such observation is relevant for cell-based clinical applications. CFD simulations for conventional radial freezing, elucidated that the ice formed at the top of the vial is of high porosity for freezing under 10% v/v DMSO, but of low porosity for lower DMSO concentrations. The simulations reveal that later conditions can result in an increase in the shear stress, up to an order of magnitude, on cells. Overall, this study provides a rational for 10% v/v DMSO to be the optimum reported concentration for conventional freezing methods, as a result of poor control ice growth direction and higher mechanical stresses at lower DMSO concentrations. Experimental results show that bottom-up freezing, using only 2.5% v/v DMSO, allow to reach cell viabilities so high as the ones obtained for 10% v/v DMSO conventional radial freezing protocols. Importantly, the results support cell cryopreservation strategies, such as bottom-up freezing, that by controlling heat transfer direction allow using lower DMSO concentrations.