In Portugal, in 2012 and 2013, the cultivation of GM maize covered 9278 ha and 8202.2 ha, respectively, representing approximately 6-7% of the total maize sowing areas. The large-scale cultivation of GM maize is not uniformly distributed along the country. It has its highest expression in the Centre-South regions. Maize is a cross-pollinating species and, preferentially, its pollen is wind-dispersed although it may also be insect-dispersed. In 2011, the European Court of Justice decided upon the need to demonstrate that pollen is a natural constituent of honey rather than an ingredient. Later, in 2013, the European Parliament defined pollen as a natural constituent of honey. This decision avoided strong financial implications concerning the need of honey labelling whenever Genetically Modified (GM) pollen makes up more than 0.9% of the species pollen fraction, according to the Regulation (EC) 1829/2003. Contrarily, the presence of flour, which is not a natural component of honey but may occur when industrial mills are in the neighborhood of the hives, might be treated according to the legal labelling framework for ingredients. Therefore, there is a need to distinguish between pollen and flour. Currently, the detection/quantification of GM components in honey is done by real-time PCR. However, until now no procedure can distinguish pollen from other kinds of GM material. Our main goal was, therefore, to develop a reliable and accurate method to allow distinguishing pollen from flour in honey. We investigated the ability of quantitative real-time PCR together with plasmid calibrants, triploid maize seed endosperms and haploid maize pollen, to develop a new approach to differentiate between adventitious presence of GM pollen and adventitious presence of GM flour. Plasmid calibrant certified for a 1:1 copy number ratio (transgene copy number in relation to an endogenous gene copy number) allow for the distinction between triploid and haploid tissues. A seven point dilution series from a 2x106 plasmid copies/μL solution was used to establish two calibration curves, being one for the transgene and other for a species-specific gene. DNA was extracted from honey, originally obtained in a Portuguese GM free zone (Região Autónoma da Madeira), spiked either with pollen, flour or embryos from a hemizygous population having the female progenitor as the transgene donor. PCR efficiencies were of 93 and 94% for both reactions. As proof-of-concept, samples of national and imported honey, commercialized either at local producers or in the supermarkets, were analyzed and the quantification results compared with the spiked honey samples. With this information available, correct estimation of the relative transgene copy number allowed for the distinction between pollen and flour. We have adapted real-time PCR to fit into the requirements of GMO labelling regulations. This approach has considerable potential to evaluate escapes of maize flour and for the establishment of recommendations for the milling companies in order to minimize effective entry of flour into the hives. The procedure has been in-house validated for maize.