BACKGROUND & AIMS: Celiac disease is characterized by HLA-DQ2/8-restricted responses of CD4+ T cells to cereal gluten proteins. A diagnosis of celiac disease based on serologic and histologic evidence requires patients to be on gluten-containing diets. The growing number of individuals adhering to a gluten-free diet (GFD) without exclusion of celiac disease complicates its detection. HLA-DQ-gluten tetramers can be used to detect gluten-specific T cells in blood of patients with celiac disease, even if they are on a GFD. We investigated whether an HLA-DQ-gluten tetramer-based assay accurately identifies patients with celiac disease. METHODS: We produced HLA-DQ-gluten tetramers and added them to peripheral blood mononuclear cells isolated from 143 HLA-DQ2.5 subjects (62 subjects with celiac disease on a GFD, 19 subjects without celiac disease on a GFD [due to self-reported gluten sensitivity], 10 subjects with celiac disease on a gluten-containing diet, and 52 presumed healthy individuals [controls]). T cells that bound HLA-DQ-gluten tetramers were quantified by flow cytometry. Laboratory tests and flow cytometry gating analyses were performed by researchers blinded to sample type, except for samples from subjects with celiac disease on a gluten-containing diet. Test precision analyses were performed using samples from 10 subjects. RESULTS: For the HLA-DQ-gluten tetramer-based assay, we combined flow-cytometry variables in a multiple regression model that identified individuals with celiac disease on a GFD with an area under the receiver operating characteristic curve value of 0.96 (95% confidence interval [CI] 0.89-1.00) vs subjects without celiac disease on a GFD. The assay detected individuals with celiac disease on a gluten-containing diet vs controls with an area under the receiver operating characteristic curve value of 0.95 (95% CI 0.90-1.00). Optimized cutoff values identified subjects with celiac disease on a GFD with 97% sensitivity (95% CI 0.92-1.00) and 95% specificity (95% CI 0.84-1.00) vs subjects without celiac disease on a GFD. The values identified subjects with celiac disease on a gluten-containing diet with 100% sensitivity (95% CI 1.00-1.00]) and 90% specificity (95% CI 0.83-0.98) vs controls. In an analysis of 4 controls with positive results from the HLA-DQ-gluten tetramer test, 2 had unrecognized celiac disease and the remaining 2 had T cells that proliferated in response to gluten antigen in vitro. CONCLUSIONS: An HLA-DQ-gluten tetramer-based assays that detects gluten-reactive T cells identifies patients with and without celiac disease with a high level of accuracy, regardless of whether the individuals are on a GFD. This test would allow individuals with suspected celiac disease to avoid gluten challenge and duodenal biopsy, but requires validation in a larger study. Clinicaltrials.gov no: NCT02442219.