AIMS: Numerous studies have suggested that transfusion of red blood cells (RBCs) stored over a long period of time may induce harmful effects due to storage-induced lesions. However, the underlying mechanisms responsible for this damage have not been identified. Furthermore, it is unclear why and how up to 30% of long-stored RBCs disappear from the circulation within 24 hours after transfusion. The aim of this study was to determine how the cell number of RBCs of different ages changes during storage and how these cells undergo cumulative structural and functional changes with storage time. METHODS AND RESULTS: We used Percoll centrifugation to fractionate the RBCs in blood bank stored RBC units into different aged sub-populations and then measured the number of intact cells in each... sub-population as well the cells' biomechanical and biochemical parameters as functions of the storage period. We found that the RBC units stored for ≤ 14 days could be separated into four fractions: the top or young cell fraction, two middle fractions, and the lower or old fraction. However, after 14 days of storage, the cell number and cellular properties declined rapidly whereby the units stored for 21 days only exhibited the three lower fractions and not the young fraction. The cell number within a unit stored for 21 days decreased by 23% compared to a fresh unit and the cells that were lost had hemolyzed into harmful membrane fragments, microparticles, and free hemoglobin. All remaining cells exhibited cellular properties similar to those of senescent cells. CONCLUSION: In RBC units stored for greater than 14 days, there were fewer intact cells with no healthy cells present, as well as harmful membrane fragments, microparticles, and free hemoglobin. Therefore, transfusion of these stored units would not likely help patients and may induce a series of clinical problems.