We recently showed that a single amino acid substitution of tryptophane into glycine at residue 167 facing the "A pocket" forms a novel HLA-B51 subtype, B*5103, which is serologically discriminated as HLA-BTA. CDC assay of human alloantisera specific for the HLA-B5 CREG against B*5103- or B*5101-transfected human B-cell line, Hmy2C1R (C1R), supported the belief that human alloantisera can discriminate B*5103 from B*5101 Ag. Moreover, we found that 4D12 anti-B5, B35 CREG mAb cannot bind to B*5103 Ag on C1R cells or L cells although it binds to B*5101 Ag on both cells. These results indicate that alloantibodies can detect a single amino acid substitution at residue 167. Furthermore, it was suggested that 4D12 mAb recognizes the structure formed by the HLA-peptide complex since this mAb did not bind to empty HLA-B5, B35 CREG Ag on RMA-S transfectants. Six of eight anti-HLA-B*5101 CTL clones are not able to kill C1R cells expressing B*5103, indicating that conformational change of the A pocket by substitution at residue 167 has a crucial influence on recognition of alloreactive T cells. Therefore, discrimination of B*5103 from B*5101 would seem to be important in bone marrow transplantation.