Studies by Thompson, 1 Carter et al, 2 MacDonald et al. and McCrae 4 have all shown a high familial frequency of coeliac disease (CD). The diagnostic criteria differed in each study; in Thompson's patients the diagnosis was based on clinical or historical data alone; in Carter's study response to a gluten-free diet was required, while in MacDonald's and McCrae's studies the diagnostic criteria included the demonstration of the lesion of CD by small bowel biopsy. In MacDonald's study some asymptomatic cases were discovered and a few of these had normal fat balance results. No hereditary pattern emerged from these studies though MacDonald postulated a dominant gene with variable penetrance while McCrae suggested that susceptibility to CD is inhereited multifactorially and that enviromental factors other than dietary gluten are of aetiological importance. Hoffman et al. 5, on the other hand, have described discordance for CD in identical twins.