**5. Overview of methods for the determination of drug residues in honey**

In the following paragraphs for each compound or class of compounds, an overview of the published confirmatory methods for the determination of residues in honey is given in **Tables 3**–**11**. Although widely applied in routine laboratories as screening methods, procedures based on bioanalytical techniques such as immunoenzymatic or receptor tests are not considered.

In **Tables 3**–**11**, for each reviewed procedure, the method limits (CCβs or LODs) are reported. Method limits are generally estimated by the LOD parameter, but, unfortunately, Commission Decision 2002/657/EC [6] introduced a different terminology, that is, decision limit (CCα) and detection capability (CCβ). Although the estimation of method limits is one of the most problematic topics of analytical chemistry [15], from a theoretical point of view, LOD and CCβ (for banned substances) are essentially the same parameter taking into account of both alpha-error (false-positive rate) and beta-error (false-negative rate) [16]. In the relevant column of the tables, CCβs are reported, if available, or, alternately, LODs. They are always given with a maximum of two significant figures.

The most used technique is LC-MS, in particular LC-QqQ platform (**Tables 3**–**1**). The need of reaching low concentrations involves a progressive decline of LC-DAD- and LC-FLDbased procedures. Methods based on LC-MS (single quadrupole) platform are sporadically described. Finally, in the last few years, high-resolution mass analysers are more and more applied. The ionisation source is almost always electrospray in positive mode (ESI+), except for chloramphenicol for which negative ionization is largely favoured (ESI−). The chromatographic separation is generally achieved in reversed-phase mode, except for aminoglycosides (streptomycin and dihydrostreptomycin) where HILIC columns are frequently applied.
