The aim of the study was to analyze the accuracy of (i) HLA-DQ typing and anti-tissue transglutaminase antibodies immunoglobulin A (tTG-IgA) serology and (ii) a "simple scoring system" (SSS) for the diagnosis of celiac disease (CD). The study included 91 patients with positive tTG-IgA, who had been tested for HLA-DQ. Patients were divided into 3 groups: typical CD, atypical CD, and non-CD. The sensitivity, specificity, positive (PPV) and negative predictive value (NPV), positive (PLR) and negative likelihood ratio (NLR) and accuracy of the test combining genotyping and tTG-IgA positivity and the simple scoring system for the diagnosis of CD were evaluated. The combination of genotyping and strong tTG-IgA positivity had a sensitivity of 93.5%, specificity of 61.5%, PPV of 93.5%, NPV of 61.5%, PLR of 2.4, NLR of 0.1 and accuracy of 89% for "CD." SSS had a higher specificity (84.6%), higher PPV (97.3%), higher NPV (68.7%), higher PLR and higher accuracy (92.3%). The combination of genotyping and strong tTG-IgA positivity missed two patients with typical CD (4%) and three patients with atypical CD (10.7%). Two cases with malabsorptive symptoms (33.3%) and three patients without malabsorptive symptoms (42.8%) would have been misdiagnosed as CD if these tests were used. Intestinal biopsy is still mandatory for diagnosis of CD in Turkish children.