The reliability and comparability of gluten analytical results in gluten-free foods is hampered by the lack of reference materials (RM). This is partly caused by the yet incomplete knowledge of the effect of genetic and environmental variability of wheat proteins on immunochemical analyses, which affects the choice of gluten source to be applied for RM production. We investigated the genetic variability and the effect of harvest year on the protein composition of five previously selected wheat cultivars. We also compared the magnitude of these effects on ELISA results to get closer to the question of choosing individual cultivar or a mixture as an RM. Our results proved that the application of a blend for this purpose is advantageous. The candidates were also produced on pilot scale to investigate the feasibility of upscaling. The results of comparison studies showed that the pilot scale blended flour can also be suitable for RM.