OBJECTIVES: The aim of the study was to describe diagnostic criteria used in children with coeliac disease (CD) and selective IgA deficiency; to determine if the publication of the 2012 ESPGHAN criteria prompted any changes; to evaluate the evolution of serological markers. METHODS: Multicenter, retrospective, descriptive study of a cohort of children under 15 years with selective IgA deficiency diagnosed with CD (January 2006 to December 2016). Demographic, clinical, genetic, histological and IgG-based antibodies were collected at diagnosis and follow-up. RESULTS: Eighty-six children were included, 60 diagnosed after the guide. Two groups were established: G1 (n = 63) and G2 (n = 23) with or without diagnostic biopsy respectively. In G1: 87.3% were symptomatic, 87.3% had human leukocyte antigan (HLA) DQ2/DQ8 typing (all positive), all had IgG serology positive (71.5% ATG, 35% EMA, 19% DPG, 9.5% AGA), and all had villous atrophy (Marsh-Oberhuber 2-3). Follow-up data were available in 58 children, 34 after 2 years on a gluten-free diet. Fifty-two percentage remained ATG IgG-positive despite good dietary adherence and symptom remission. Regarding G2: all were diagnosed post-2012, had typical symptoms, HLA DQ2/DQ8 positive and ATG IgG × 10 ULN. Additionally, EMA IgG was performed in 14 (60%), all positive. CONCLUSIONS: In our cohort of children with selective IgA deficiency and diagnosed with CD, children without a diagnostic biopsy suggests that IgG serology was considered the equivalent as IgA isotype, even when this is not addressed in the aforementioned guidelines. Great heterogeneity was observed in the IgG serology used at diagnosis. After 2 years of a gluten-free diet, half of children remained with a positive serology.