© 2017 Catozzi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. The aim of this study was to define the microbiota of water buffalo milk during sub-clinical and clinical mastitis, as compared to healthy status, by using high-throughput sequencing of the 16S rRNA gene. A total of 137 quarter samples were included in the experimental design: 27 samples derived from healthy, culture negative quarters, with a Somatic Cell Count (SCC) of less than 200,000 cells/ml; 27 samples from quarters with clinical mastitis; 83 samples were collected from quarters with subclinical mastitis, with a SCC number greater of 200,000 cells/ml and/or culture positive for udder pathogens, without clinical signs of mastitis. Bacterial DNA was purified and the 16S rRNA genes were individually amplified and sequenced. Significant differences were found in milk samples from healthy quarters and those with sub-clinical and clinical mastitis. The microbiota diversity of milk from healthy quarters was richer as compared to samples with sub-clinical mastitis, whose microbiota diversity was in turn richer as compared to those from clinical mastitis. The core microbiota of water buffalo milk, defined as the asset of microorganisms shared by all healthy milk samples, includes 15 genera, namely Micrococcus, Propionibacterium, 5-7N15, Solibacillus, Staphylococcus, Aerococcus, Facklamia, Trichococcus, Turicibacter, 02d06, SMB53, Clostridium, Acinetobacter, Psychrobacter and Pseudomonas. Only two genera (Acinetobacter and Pseudomonas) were present in all the samples from sub-clinical mastitis, and no genus was shared across all in clinical mastitis milk samples. The presence of mastitis was found to be related to the change in the relative abundance of genera, such as Psychrobacter, whose relative abundance decreased from 16.26% in the milk samples from healthy quarters to 3.2% in clinical mastitis. Other genera, such as SMB53 and Solibacillus, were decreased as well. Discriminant analysis presents the evidence that the microbial community of healthy and clinical mastitis could be discriminated on the background of their microbiota profiles.