Background: According to national guidelines or directives, monoclonal ABO reagents may be required to detect Ax and B weak subgroup red blood cells (RBCs). Many routine laboratories do not have access to naturallyoccurring ABO subgroups that can be used as weak controls for these reagents. Group O RBCs modified with synthetic analogs of blood group A and/or B glycolipids (KODE technology) to mimic weak ABO subgroups could be used for quality control purposes. Aim: Extensive serological testing of KODE RBCs has previously been performed. An extended evaluation of KODE RBCs using flow cytometry was performed to explore the correlation between the concentrations of synthetic glycolipids and A/B site density of the resulting RBCs. The aim of this study was to examine if KODE RBCs mimic the distinct flow cytometric patterns of naturally-occurring ABO subgroups. Material and Methods: KODE RBCs were prepared according to a previously decribed procedure [Frame et al., Transfusion 2007; 47: 876–82]. RBCs were modified with 15 different concentrations of synthetic glycolipids, ranging from 1 mg/mL to 60 ng/mL for KODE-A and 5 mg/mL to 0.3 lg/mL for KODE-B. The concentration was decreased by doubling dilution steps. Sensitive and specific flow cytometry [Hult & Olsson. Transfusion 2006; 9S: 32A] was used to characterize and semiquantify the synthetic A and B antigen levels on RBCs. Relevant control RBCs (A1, A2, Ax, B, Bweak and O) were included in each run. For both KODE-A and KODE-B RBCs, repeat samples were produced for four selected concentrations and all KODE batches were tested in triplicate. Results: Flow cytometric testing of KODE RBCs modified with high concentrations of synthetic glycolipids revealed a uniform and even distribution of antigens in the cell population as shown by a single narrow peak in the FACS histograms. When lower concentrations were used, peaks tended to broaden to a pattern found in Ax and most B subgroups indicating a more variable antigen site density on the cells in the population. The concentrations of synthetic glycolipids that produced KODE cells that resembled the naturally-occurring subgroup control RBCs used in this study are ~2–4 lg/mL for KODE-A and ~10 lg/mL for KODEB. Repeat testing demonstrated good correlation between flow cytometric runs. Discussion and Conclusion: Using very low amounts of synthetic glycolipids, KODE-A and KODE-B RBCs can be made to mimic Ax and Bweak subgroup control RBCs, respectively, according to this flow cytometry method. With higher concentrations of synthetic glycolipids, the KODE RBCs demonstrated a more uniform and even distribution of antigens among the cells. This is in contrast to naturally-occurring subgroups in which some cells express almost no A or B antigen whilst others have close to normal levels. The reason for this is unknown. KODE RBCs obviously lack A carrying glycoproteins but it is not fully understood to what extent glycolipid versus glycoprotein epitopes contribute to the phenotype of weak subgroups. This study indicates that KODE RBCs with weak expression of A and/or B antigen have characteristics compatible with use as quality controls for monoclonal ABO reagents and could be a valuable addition in the serological laboratory.