Objective: Tissue-transglutaminase antibodies (TGA) may be used to diagnose celiac disease (CD) without biopsy in selected cases. We aimed to investigate real-life performance of a CD serology automated analyzer (Bioplex2200), and to explore the correlation between TGA levels and intestinal biopsies in children. Methods: A retrospective review was performed in 2 pediatric gastroenterological centers, between November 1, 2018 and April 1, 2020 and included patients with both TGA serology testing and duodenal biopsies. Retrieved data included patients' demographics, medical background, TGA levels, and biopsy results. Results: Overall, 538 children were evaluated, 256 with positive TGA (68.4% girls, median age 6.4 years), and 282 with negative TGA (53.9% girls, median age 13.4 years). Among patients with positive TGA, intestinal biopsies confirmed CD in 219 (85.5%). Overall, positive serology with normal histology was found in 14.5% of the cohort, with 52%; 21.6%; 21.1%; and 4.2% in TGA ranges of 1 to 3 times upper limit of normal (ULN); 3 to 5 ULN; 5 to 10 ULN; and above 10 times ULN, respectively, P < 0.001. Area under the receiver-operating characteristic curve (AUC) was 0.963 (95% CI 0.947-0.980). Among patients with positive TGA, 216 (84.4%) had positive antiendomysial antibodies. In this sub-group, the overall diagnostic performance was inferior, with AUC of 0.737 (95% CI 0.834-0.839). Conclusions: The Multiplex TGA assay had a very high diagnostic accuracy in real-life. Among patients with positive TGA, adding EMA did not improve the diagnostic performance of the test. False-positive rates differed between different ranges of TGA and were low with TGA above 10 times ULN.