The HIV-1 Gag p24 protein contains the HLA class-1 B*2705 restricted epitope KK10, responses to which are associated with delayed progression. Data from in vitro proteasomal digestion studies from our group has shown the production of a number of C-terminally extended and truncated epitopes containing KK10, produced in far higher quantities during proteasomal digestion than this “optimal epitope” and that the amount of antigen made in proteasomal digestion is instrumental in determining the development of immunodominance. This work aims to characterise the contribution of these naturally processed epitope forms to the cellular immune response to this region. Further proteasomal digestion studies have shown that the common KK10 intra-epitope escape mutant sequences R132K and L136M have major effects on epitope production by the proteasome and that a range of short peptides containing the N-terminal of the KK10 sequence are produced in large quantities by the proteasome. Recognition of the KK10 epitope forms by HLA B*2705 HIV-1 patients were characterised ex vivo and show recognition of KK10 epitope forms somewhat independent of the presence of KK10 recognition, we also show cross-recognition between KK10 epitope forms by CD8+ T cells, as well as recognition by CD4+ T-cells. TCR from CD8+ T-cells specific for KK10 epitope forms were found to share common features in the HLA binding CDR hyper-variable loops. Structural studies of the HLA B*2705 molecules in complex with the KK10 epitope forms show a shared binding motif at the N-terminus, and to a lesser extent, the C-terminus of the binding groove which may facilitate cross-recognition of complexes. In addition these studies show a potentially novel binding mode for a 14mer peptide, and refolding of truncated KK10 peptides as short as a 4mer with the HLA B*2705 molecule (crystallisation with a 6mer peptide shown). This demonstrates previously unrecognised flexibility of the HLA class-1 to bind and present peptides of different lengths to T-cells. We show that these HLA B*2705 binding-capable truncated peptides do not induce a CD8+ T-cell response in HLA B*2705 HIV-1 patients and may be able to block CD8+ T-cell responses to the KK10 epitope. This might represent a novel form of viral CTL escape. In addition we observe the presence of KK10 flanking mutations in patient sequences and significant associations between the presence of intra-epitope escape mutations, KK10 recognition and patterns of escape in flanking sequences. Finally we note the reduction in binding of KIR3DL1 to KK10 epitope forms relative to the KK10 epitope-HLA B*2705 complex. The presence of HLA B*2705 and KIR3DL1 associate with improved disease course in HIV-1 though the mechanism through which this occurs has yet to be defined.