. The DNA was prepared and amplified fragment length polymorphism (AFLP) genotyping was done on the final CNS collection as described previously (Piessens et al., 2010). The BioNumerics software version 6.01 (Applied Maths, Sint-Martens-Latem, Belgium) was used for normalization of fingerprints, and the library and identification module was used for calculation of genetic similarities and identification of the field isolates. Species identification was done based on similarity of AFLP fingerprints to entries in the staphylococcal AFLP library, which contained fingerprints of 54 CNS type and reference strains, representing 49 different CNS species and subspecies, and 247 well-identified bovine CNS isolates belonging to 18 CNS species common in cattle (Piessens et al., 2010). Similarities were calculated based on the Pearson product-moment correlation coefficient, and 50% similarity to a library entry was used as a cut-off for species identification. When an isolate showed less than 50% similarity to all library entries in the numerical analysis, its fingerprint was visually compared with library strains in a dendrogram constructed with the unweighted pair group method with arithmetic mean (UPGMA) clustering algorithm. Due to different intensities of fingerprints, similarity can be