OBJECTIVES: There is currently much call for a reliable enzyme-linked immunosorbent assay (ELISA) protocol for determining gluten in foods to serve as a basis for further Codex Alimentarius regulations. Given its ability to recognize the potential coeliac-toxic epitope QQPFP, which occurs repeatedly in alpha-, gamma- and omega-gliadins, hordeins and secalins, the monoclonal antibody R5 raised against a secalin extract may prove to be an essential tool for gluten analysis. This study was designed to develop a highly sensitive and specific sandwich ELISA to quantify low levels of wheat, barley and rye prolamins in foods for coeliacs. METHODS: Simple sandwich ELISA based on the use of a single monoclonal antibody (R5) as both the coating and detection was developed. A quantitative cocktail gluten-extraction procedure for heat-processed foods was also tested. RESULTS: R5-ELISA was able to identify gliadins, hordeins and secalins with assay sensitivities of 0.78, 0.39 and 0.39 ng/ml, respectively. The assay's detection limit was 1.5 ng gliadins/ml (1.56 ppm gliadins, 3.2 ppm gluten). The system proved insensitive to the non-coeliac-toxic cereals maize, rice and oats, and was non-cultivar-dependent. It was also able to detect gliadins and hordeins in unprocessed and heat-processed wheat- and barley-based products, and to estimate the gluten content of hydrolysed foods. CONCLUSION: We present a new generation of a robust sandwich R5-ELISA with good reproducibility (8.7%) and repeatability (7.7%). Its gluten-detection limit of 3.2 ppm is lower than the existing threshold of 20-200 ppm. The ELISA, which is equally sensitive to barley, wheat and rye prolamins, is compatible with the quantitative cocktail extraction procedure for heat-processed foods. Along with the cocktail procedure, the Working Group on Prolamin Analysis and Toxicity is currently evaluating an R5-ELISA system as proposed by the Codex Alimentarius Commission.