In special dietary products for people intolerant to gluten, gluten content is not supposed to exceed the regulatory thresholds. Enzyme-linked immunosorbent assays (ELISAs) are routinely used to quantitate gluten in these products. They measure gliadin/gluten with high specificity and sensitivity, but they have some limitations. Quantitative and qualitative variability of the target proteins among wheat cultivars is a factor that may cause inaccurate results. One of the aims of this work was to characterize the protein composition of five wheat cultivars grown in multiple harvest years and their blends by reversed-phase high-performance liquid chromatography (RP-HPLC). The gliadin/gluten content of these wheat flours was also analysed with two commercial ELISA kits. The effect of differences in protein profiles between the flours from an individual cultivar and the blend of five cultivars, harvest years, as well as processing procedures (dough forming and baking) on the results of two ELISA kits was investigated and their relative magnitude was determined. Among the factors investigated, the differences between flours had the greatest impact on gliadin recoveries.