Background:In order to assess their long-term predictability for the diagnosis of celiac disease, antiendomysium antibody results were compared with the outcome of the Interlaken diagnostic process.Methods:Prospective gluten challenge was performed in 153 children with previously diagnosed flat small-intestine mucosa. In 90 patients (Group A), endomysium antibodies were initially positive, in seven (Group B) they were negative, and 56 patients (Group C) had no initial serological results. In IgA-deficient persons, IgG antibodies were also assayed, both by the immunofluorescent method.Results:Histological relapse rates were 100% (90/90), 14.3% (1/7), and 76.8% (43/56), p < 0.001, in Groups A, B, and C, respectively. Each patient with relapse also exhibited endomysium antibody positivity during the challenge. Patients in whom celiac disease could be finally ruled out remained consistently endomysium-antibody negative. The celiac disease patient in Group B had severe secondary immunoglobulin deficiency at entry, which explained the initial negativity. Diagnosis based on antiendomysium antibody positivity and flat mucosa gave a higher applicability (92.8 vs. 50.3%) and reliability (relapse rate 100 vs. 89.6%) than the 1990 European Society of Paediatric Gastroenterology and Nutrition (ESPGAN) criteria among these patients.Conclusions:Endomysium antibody positivity at presentation has been found to be as useful as gluten challenge in the diagnosis of celiac disease, even in patients under the age of 2 years. Challenge is still advisable in patients with a flat small intestinal mucosa when antiendomysium antibody results are negative or have not been done, as among these patients significantly lower relapse rates were found.