Extraintestinal manifestations are commonly seen in (up to 40%) in patients with either ulcerative colitis (UC) or Crohn’s disease (CD). Many of them are associated with active disease (recurrent mouth ulcers, erythema nodosum, uveitis, Type 1, or reactive large joint arthropathy). Others pursue a clinical course independent of the intestinal disease (sacroiliitis, ankylosing spondylitis, Type 2 or small joint arthropathy, primary sclerosing cholangitis). Pyoderma gangrenosum usually occurs with active intestinal disease but can occasionally occur when the intestinal inflammation is largely in remission. For CD, both groups of these extraintestinal manifestations occur in patients with colonic involvement, and rarely occur when only small intestinal disease is present. Data on the HLA associations are limited both in terms of numbers of studies and numbers of patients studied. However, a particularly strong association has been found between the Class II allele DRB1*0103, and Type 1 arthropathy.1 Of patients with either UC or CD who develop the acute, large joint arthropathy in association with active disease, 38% of them carry this allele, compared with about 8% of UC patients as a whole, and only 2% of the healthy population. It is interesting to note that patients who develop a reactive arthropathy following infection (Salmonella, Campylobacter, Reiter’s syndrome) also have a high carriage rate of this allele. This allele is also associated with the risk of having recurrent mouth ulcers or uveitis, although less strongly compared with Type 1 arthropathy. Uveitis and recurrent mouth ulcers are also associated with HLA B27. Erythema nodosum has a particularly strong association with a polymorphism in the promoter region of the TNF gene (-1031C).2 There are further HLA associations with some of those extraintestinal manifestations that are independent of inflammatory bowel disease (IBD) activity. Ankylosing spondylitis (AS) occurs in 1%–2% of patients with IBD and differs from sporadic AS in having no gender association (compared with a marked male predominance in sporadic AS) and in having a weaker association with HLA B27 (60% compared with 90% in sporadic AS). Nevertheless, the association is highly significant. Patients with sacroiliitis only have a 22% risk of having HLA B27, a percentage that is little different from the normal control population. Patients with Type II arthropathy, a seronegative, symmetrical small joint disease, have a strong association with HLA B44. Data concerning primary sclerosing cholangitis (PSC) are conflicting. Most studies find an association with HLA DR3. In the largest study the haplotype HLA DRB1*0301, DQB1*0201 (the classical autoimmune haplotype) was found in 25% of patients with PSC and colitis, 16% of patients with pancolitis only, and 10% of controls (Cullen S, Chapmann RG, Jewell DP. Submitted). These data clearly indicate that individuals possessing certain HLA alleles are at risk of developing extraintestinal manifestations, should they have the misfortune of developing IBD. In contrast, they are unlikely to occur in patients with IBD who do not possess these alleles. Furthermore, these associations can explain why these manifestations often overlap in a single individual. The outstanding questions concern the mechanisms of pathogenesis. Since the HLA region on the short arm of chromosome 6 is a very gene-dense region, it cannot be assumed that the allelic associations that have been identified are the relevant genes as opposed to having identified an allele which is in linkage disequilibrium with a more relevant polymorphism in a neighboring gene. Until this problem has been overcome, functional studies to explore pathogenic mechanisms cannot sensibly begin.