The dual observations that human leukocyte antigens have an antigen-binding groove and that the polymorphism we study as human leukocyte antigen types is largely related to amino acid substitutions in and around that groove have provided a new focus for immunogenetic studies. In autoimmune liver disease, recent studies have described specific amino acid substitutions in the antigen-binding groove of human leukocyte antigen DR molecules that may determine both disease susceptibility, through their direct influence on antigen binding, and the severity of the disease. In autoimmune hepatitis, lysine residues at DR beta position 71 in European subjects and arginine or histidine residues at DR beta position 13 in Japanese subjects may be responsible for much human leukocyte antigen-encoded disease susceptibility. Similar claims have been made for leucine residues at DR beta 38 in primary sclerosing cholangitis and for leucine residues at DP beta 35 in Japanese patients with primary biliary cirrhosis. To date, our knowledge of genetic susceptibility to autoimmune liver disease is incomplete. Other genes may contribute to susceptibility to autoimmune liver disease--for example the contribution of TAP genes, upstream promoter sequences and class III genes on chromosome 6 and the T-cell receptor genes and complement genes elsewhere in the human genome is currently unclear. Additional information concerning the immunogenetic contribution to disease severity is needed to complete the picture.