The role of the bound peptide in alloreactive T-cell recognition is controversial, ranging from peptide-independent to peptide-specific recognition of alloreactive T-cells. The aim of this study is to find the evidence that there exist peptide/MHC complex (pMHC)-specific CTLs among alloreactive T cells generated with long-term mixed lymphocytes culture (LTMLC). A single pMHC was manipulated by loading the TAP-defective, HLA-A2 expressing T2 cells with a viral peptide (LMP2A(426-434)) or a self-peptide (Tyr(369-377)). The PBLs samples from 4 HLA-A2 positive (HLA-A2+ve) and 4 HLA-A2 negative (HLA-A2-ve) donors were included in this study. The HLA-A2+ve PBL co-cultured with the LMP2A(426-434) pulsed T2 (T2/LMP) stands for the nominal T-cell response to a viral antigen, and the HLA-A2-ve PBLs co-cultured with the Tyr(369-377) pulsed T2 (T2/Tyr) for alloreactive T-cell response to an allogeneic antigen. The specificity of the expanded CTLs after the LTMLC was detected by their specific cytotoxicity and binding ability to specific pMHC-tetramer. An HLA-A2 restricted, HIV peptide (Gag(77-85)) was included for control. The cultural bulk of HLA-A2+ve PBLs with the T2/LMP showed an elevated specific cytotoxicity against the T2/LMP compared to that against the T2/HIV (26.52%+/-3.72% vs 7.01%+/-0.87%, P<0.001), and an increased frequency of binding to LMP-tetramer compared to that binding to HIV-tetramer (0.98%+/-0.33% vs 0.05%+/-0.01%, P=0.0014). The cultural bulk of HLA-A2-ve PBLs with the T2/Tyr showed a more active cytotoxicity against the T2/Tyr than that against T2/HIV (28.07%+/-2.58% vs 6.87%+/-1.01%, P<0.001), and a higher frequency of binding to the Tyr-tetramer than that binding to the HIV-tetramer (0.88%+/-0.3% vs 0.06%+/-0.03%, P=0.0018). Our results indicate that the LTMLC is able to expand the viral antigen-specific CTLs as well as allogeneic antigen-specific CTLs. A relatively large proportion of alloreactive CTLs should be pMHC-specific, i.e., the specificity of the alloreactive lines depends on both the bound peptide and the allotype of MHC. Our observations support the hypothesis that the cumulative effect of T cells specific to each peptide epitope could account for the strength and diversity of the alloresponse. The method using manipulated pMHC and the LTMLC to generate pMHC-specific, alloreactive CTLs is of potential importance for adoptive T-cell immunotherapy.