SUMMARY We investigated the incidence of circulating corneal epithelium antibodies in patients with uveitis. A high percentage of the patients (42%) were positive, whereas only 4% of controls had these antibodies in their serum. Significantly more patients with anterior and diffuse uveitis had corneal epithelium antibodies than did those with posterior uveitis. Subdivision of anterior uveitis into HLA-B27 positive versus negative patients showed a higher incidence of the antibodies in the HLA-B27 positive group. A previous history of uveitis may play a role in the pathogenesis of peripheral corneal thinning diseases, for which an autoimmune aetiology has been suggested.The peripheral corneal thinning syndrome includes many corneal ulcerative diseases such as Mooren's ulcer, Wegener's granulomatosis, polyarteritis nodosa, Terrien's disease, and rheumatoid arthritis. Many possible causes for this syndrome have been proposed, but as yet none have been proved. A recent study' showed that circulating immune complexes may play a role in the pathogenesis of Mooren's ulcer and marginal ulceration in the presence of collagen vascular disease. However, earlier studies suggested a humoral autoimmune response against corneal antigens as a pathogenetic mechanism of the Mooren's ulcer. ' In 1969 Schaap et al.2 were the first to show the presence of circulating antibodies to corneal epithelium in a patient with Mooren's ulcer. This finding was confirmed by Brown and Mondino34 and other workers.56 In a recent study from our institute Van der Gaag et al.' found a high incidence of circulating corneal epithelium antibodies (81%) in African patients with a Mooren's ulcer. A local control group, however, also had a high prevalence (87%) of these antibodies.These findings raised the question whether the presence of circulating antibodies to corneal epithelium is the primary cause of the ulceration or whether these antibodies reflect a secondary phenomenon following a generalised parasitic or viral infection.8 Another hypothesis may be that these Correspondence to P J Kruit, MD,