This part of ISO 6579 gives a method for the enumeration of Salmonella spp. present in: — products intended for human consumption and for the feeding of animals; — environmental samples in the area of food production and food handling; — animal faeces; — environmental samples from the primary production stage; by calculation of the most probable number (MPN). The method is based on miniaturization of the dilution, pre-enrichment and selective enrichment steps. The selective enrichment medium, modified semi-solid Rappaport–Vassiliadis (MSRV), is intended for the detection of motile salmonellae and is not appropriate for the detection of non-motile salmonellae. It is possible that the method is less appropriate to enumerate Salmonella ser. Typhi and Salmonella ser. Paratyphi. The method is not appropriate for the enumeration of Salmonella spp. in (very) low contaminated samples (<1 cfu/g). NOTE The number of non-motile salmonellae is generally low in most of the matrices relevant for this method. In this note, examples are given for samples from primary production. The non-motile Salmonella biovars of Salmonella Gallinarum (Salmonella Gallinarum biovar gallinarum and Salmonella Gallinarum biovar pullorum) do not seem to survive long in environmental samples and are therefore rarely detected in faecal or environmental (such as dust) samples (regardless of the method). The number of other non-motile Salmonella serovars in faecal samples seems to be generally low. For example, in Reference  in which approximately 1 000 faecal samples of poultry layer flocks and approximately 900 faecal samples of broiler flocks were analysed, less than 1 % of the total number of samples were positive in a selective broth and at the same time negative on MSRV medium (and likely to be non-motile). Similar results were found in a Dutch study with ca 3 200 faecal samples of pigs (unpublished data). On the other hand, in the case of the study reported in Reference , up to almost 40 % of positive samples would not have been detected (i.e. false negatives) if only a selective broth (in this case Rappaport–Vassiliadis) had been used instead of a semi-solid medium.