The present study confirms that, in the case of presence of over-represented pollen in honeys, pollen analysis alone cannot give trustworthy results for the determination of the botanical origin, even though their exclusion during pollen analysis, when they are present in percentages of up to 30%, could provide more accurate results. Consequently, pollen analysis should also be combined with the other analyses, especially in honeys with under-represented and over-represented pollens, to give safer results for the botanical characterisation of honeys. © 2017 Society of Chemical Industry.