Some years ago, we used our data for class I genes, proteins and peptide-binding specificities to develop the hypothesis that the chicken B-F/B-L region represents a "minimal essential MHC". In this view, the B locus contains the classical (highly expressed and polymorphic) class I alpha and class II beta multigene families, which are reduced to one or two members, with many other genes moved away or deleted from the chicken genome altogether. We found that a single dominantly expressed class I gene determines the immune response to certain infectious pathogens, due to peptide-binding specificity and cell-surface expression level. This stands in stark contrast to well-studied mammals like humans and mice, in which every haplotype is more-or-less responsive to every pathogen and vaccine, presumably due to the multigene family of MHC molecules present. In order to approach the basis for a single dominantly expressed class I molecule, we have sequenced a portion of the B complex and examined the location and polymorphism of the class I (B-F) alpha, TAP and class II (B-L) beta genes. The region is remarkably compact and simple, with many of the genes expected from the MHC of mammals absent, including LMP, class II alpha and DO genes as well as most class III region genes. However, unexpected genes were present, including tapasin and putative natural killer receptor genes. The region is also organised differently from mammals, with the TAPs in between the class I genes, the tapasin gene in between the class II (B-L) beta genes, and the C4 gene outside of the class I alpha and class II beta genes. The close proximity of TAP and class I alpha genes leads to the possibility of co-evolution, which can drive the use of a single dominantly expressed class I molecule with peptide-binding specificity like the TAP molecule. There is also a single dominantly expressed class II beta gene, but the reason for this is not yet clear. Finally, the presence of the C4 gene outside of the classical class I alpha and class II beta genes suggests the possibility that this organisation was ancestral, although a number of models of organisation and evolution are still possible, given the presence of the Rfp-Y region with non-classical class I alpha and class II beta genes as well as the presence of multigene families of B-G and rRNA genes.