Coagulase-negative staphylococci (CNS) are the main pathogens causing subclinical chronic mastitis (SCM) in dairy animals, and are a key challenge for improving milk quality. CNS are also an emerging class of nosocomial pathogens in humans. Identification of CNS species by classical phenotypic bacteriological methods is often unreliable, especially regarding animal isolates. Single-locus based phylogeny is frequently contradictory, depending on the gene studied. In this work, the CNS pan-genome was analyzed to study the taxonomy, phylogeny and host-adaptation of the prevalent CNS-species causing SCM in dairy animals in Israel. CNS isolates (n=387) were collected from SCM in different herds and regions. Preliminary bacterial species were identified by the Analytical Profile Index (API Staph ID32, Biomerieux), based on which 200 representative isolates of different CNS-species, hosts and farms were selected for whole-genome sequencing (WGS). Genomes were sequenced in Miseq with 2x250 PE libraries to about 50x average depth, assembled de novo and annotated. Genes were clustered using different similarity cut-offs, aiming to maximize the core-genome size. The main CNS-species found were S. chromogenes, S. haemolyticus, S. simulans, S. intermedius, S. epidermidis, S. xylosus, and S. caprae (in goats). At 70% similarity, the core-genome comprised 684 genes, and the pan-genome included 16,917 genes overall. Alignment of core-genome genes showed clear CNS-species specific clades, which corresponded to genomes clustering based on accessory-genes presence/absence. Forty percent of the isolates could not be reliably assigned to a species using API. These were identified by WGS, mostly as S. chromogenes and S. epidermidis. S. chromogenes showed the most variable accessory-genome. This was also the most common species to the three animal types, with isolates from sheep and goats comprising a sub-cluster, thus suggesting host-adaptation.