The method makes it possible to detect and identify spiramycin, tylosin and virginiamycin in animal feeding stuffs (feed raw materials of mainly plant origin and compound feeds) excluding mineral feeds and premixtures. The limit of detection is about 2 mg/kg for spiramycin, 1 mg/kg for tylosin and 1 mg/kg for virginiamycin. In some milk replacers, it can be slightly higher than 1 mg/kg for virginiamycin. Reported limits of detection are probably little overestimated but were fully validated during the collaborative study (see Annex B). In each laboratory, each day of analysis, spiked blank samples at 1 mg/kg for spiramycin and virginiamycin and at 0,5 mg/kg for tylosin are analysed for checking lower detection limits (see 9.2 and 9.3). These lower limits of detection are achievable, but should be established with an in-house validation first. Some other antibiotics can interfere in the detection of these 3 specific macrolide antibiotics. The known interferences are specified in Annex A of the method. That method should be used as a qualitative screening and/or a post-screening method (after microbiological plate test, for example). The follow-up of the antibiotics presence may be done by other analytical technics (LC and/or LC-MS technics) ([4], [10]). For confirmatory purposes, LCMS is required.