Background and Objectives Frozen‐thawed red blood cells (FTRCs) are useful blood components to patients with rare blood phenotypes. However, frozen red blood cells (FRCs) sometimes cause significant haemolysis after thawing due to the freeze/thaw process. In this study, we aimed to focus on the former process and reduce process‐related haemolysis. Materials and Methods Five‐day‐old red blood cells (RBCs) (5D) or 9‐week‐old RBCs (9 W) were glycerolized, pooled and split into two aliquots. RBCs were frozen using either the programmed freezer (PF) method or the deep freezer (DF) method. After 4–8 weeks, the FRCs were thawed and washed. In vitro characteristics were compared between the PF and DF methods. Nine week were used as a starting material for FTRCs with the assumption that they can mimic disqualified FTRCs with respect to Hb recovery. Results The PF method resulted in a significantly higher Hb recovery rate than the DF method (5D: 85.9 ± 2.1 vs. 81.1% ± 3.5%, p < 0.001) (9 W: 56.8 ± 4.0 vs. 52.4% ± 3.5%, p < 0.001). Both 5D and 9W‐derived FTRCs immediately after preparation prepared by the PF method were more resistible to haemolysis than those prepared by the DF method. On the other hand, there were no significant differences between PF and DF methods in Adenosine 5′‐triphosphate (ATP) and 2,3‐diphosphoglycerate (2,3‐DPG). Conclusion The PF method was more suitable for RBC freezing than the DF method in terms of Hb recovery in FTRCs. Although it was only 4%–5%, the improvement in the Hb recovery rate will contribute to a more stable supply.