Objectives: Mastitis is the most expensive disease in dairy cows. It is of multicausal etiology as the result of interactions among microorganisms, the host and the environment. The most important pathogens of bovine mastitis are Streptococcus, Staphylococcus, Escherichia coli and the other coliform microorganisms. Numerous Streptococcus species are important pathogens in animals and humans. Among the streptococci that are the main pathogens of bovine hosts are species such as S. agalactiae, S. uberis and S. dysgalactiae, which are also the most common mastitis pathogens in dairy cows. In the most clinical laboratories, the identification of streptococci is based on biochemical tests, and. in this way, up to 10-15% of isolated strains are misidentified. The strains within a particular species may differ in the same property, whereas the same strain may show biochemical variability. Since there is no classification system ompletely accurate, the development and application of genetic analysis methods improved the identification of streptococci. The aim of this research was to establish a detailed insight into the presence of certain types of streptococci in the etiology of the mastitis in dairy cows, to recognize public health significance of mammary gland infections caused by streptococci and to determine compliance of the results obtained by the classic microbiological analyses with those obtained by identification of the causative agents using molecular methods. Materials and Methods: The samples taken for bacteriological culture, prior regular milking, were collected aseptically in sterile 10 mL tubes, without additives, according to the National Mastitis Council recommendations and kept at 4 oC during transport. The samples were analyzed within 12 hours following the collection. The described microbiological testing was performed in accordance with generally approved recommendations as described in the Laboratory Handbook on Bovine Mastitis. The primary streptoccci isolation and identification were performed using aesculin blood agar. After identification, the strains were stored in glycerol broth at the temperature of - 20 oC. The frozen strains were revitalized by repeated inoculations on aesculin blood agar. During storage, the vitality of the strains was examined using random selection method. Isolation of DNA from the selected strains for species confirmation by genes detection and sequencing were carried using the so-called “cultural” isolation method. The DNA for further analyses was isolated with commercial kits and automated device. For molecular identification of Streptococcus sp. was used method based on gene sequencing and coding for 16S RNA, and comparison of sequences with sequences in the available databases. The primers used for molecular identification of species were pA and pH. The obtained amplification products were sequenced using the Sanger method in Macrogen Europe B. V., Amsterdam, The Netherland. The obtained sequences were compared with the sequences listed in the internationally available database NCBI (National Center for Biotechnology Information) and RDP (Ribosomal database Project) througu BioNumeric 7.6.3 software. The multilocus sequence typing (MLST) is a method based on determining DNA fragment sequences of approximately 450 bp size in seven housekeeping genes of S. agalactiae (adhP, pheS, atr, glnA, sdhA, glcK and tkt), S. uberis (gki, recP, ddl, tdk, arcC, tpi, and yqiL), S. canis (gki, gtr, murl, mutS, recP, Xptgc and yqiZ) and S. dysgalactiae (gki, gtr, murl, mutS, recP, xpt and atoB). Obtained type of sequences, i.e. allele profiles represent the SequenceType (ST). The obtained amplification products were sequenced using the Sanger method in Macrogen Europe B. V., Amsterdam, Netherland. The obtained nucleotide sequences were aligned, verified and compared using the BioNumerics software (version 7.6; applied Maths, Kortrijk, Belgium). Allele and ST are determined using Streptococcus MLST http://pubmlst.org/ database that can be loaded and used via BioNumerics software. Each species of Streptococcus sp. has a separate database. The combination of alleles on each locus consists of seven-member numerical code in which we compared each other based on the categorical coefficient and unweighted pair group method with arithmetic mean. Results: In the study, 54 strains of bacteria isolated from milk samples from udder quarters of dairy cows with subclinical mastitis were examined using molecular methods. By conventional microbiological examination, the strains were identified to the species (S. agalactiae, S. dysgalactiae i S. uberis) or to the genus (Streptococcus spp.) without final identification of the species. Out of 54 examined bacterial strains, 47 of them were confirmed to belong to the genus Streptococcus spp., 6 strains of Streptococcus agalactiae, 10 strains of Streptococcus dysgalactiae, 2 of Streptococcus canis and 31 strain of Streptococcus uberis, while 3 strains belonged to Enterococcus faecalis and 4 strains to the species Lactococcus lactis. All 47 Streptococcus sp. strains were typed using the MLST technique. The MLST is a highly discriminating technique because it is based on typing sequences of approximately 500 bp size within fragments of 7 so-called „house-keeping“ genes. Each different sequence of genes represents another allele. The combination of alleles to seven loci is an allele profile or ST. Since it is possible to have a large number of alleles in each locus, the samples do not have the same profile randomly. In this research we have studied four different species of Streptococcus sp. and a different set of loci was used for each species. In typed samples of S. agalactiae 4 different allele profiles were defined. The most common ST was ST314. The other STs were represented in the samples with one sample for each allelic profile. Allele profiles were examined at 7 loci: adhP, pheS, atr, glnA, sdhA, glcK, and tkt. The most polymorphic loci were atr and glcK with the Hunter-Gaston discriminatory index (HGDI) of 0.73. It is followed by adhP, glnA, sdhA and tkt with the HGDI index of 0.60. The least polymorphic was pheS with the HGDI index of 0.33. Using the eBURST algorithm, we defined that there were no allelic profiles form a group, i. e. that they were "singletons". No local distribution of allelic profiles was visible from the Minimum spanning tree (MST) view. In the samples of S. dygalactiae were defined 7 different allele profiles. The most common ST was ST305. Allelic profiles were examined at 7 loci: gki, gtr, murI, mutS, recP, xpt, and atoB.The most polymorphic locus was mutS with the HGDI index of 0.86. It is followed by gki and xpt with the HGDI index of 0.71 and 0.64, respectively, and gtr, recP and atoB with an index of 0.46. The least polymorphic was murI with the HGDI index of 0.25. Using the eBURST algorythm, we have defined only one cluster to which ST305 and ST453 belong. The regional grouping of the samples from our research was clearly visible from the MST presentation. In the typed S. canis samples, we have defined one allelic profile. An identical ST9 allele profile was determined for the two isolated strains. Allelic profiles were examined at 7 loci: gki, gtr, murI, mutS, recP, xpt, and yqiZ. The HGDI at all loci was equal to 0 because there is no polymorphism since the same ST was identified on two strains. There is no grouping using the eBURST algorythm. From the MST presentation it is clear that Streptococcus canis strains identified in our research belong to one of the most represented groups in the Europe, ST9 which very likely, forms a clonal cluster in which ST9 is even a carrier. In the S. uberis samples, we have defined 22 different allelic profiles. All allelic types have so far been unknown in the database. The most common ST was ST1203 with 9 samples. ST1210 was determined in 2 samples while the other STs were represented with one sample for each allelic profile. The allelic profiles of S. uberis were examined at 7 loci: arcC, ddl, gki, recP, tdk, tpi and yqiL. The most polymorphic loci were tdk and arcC with the the the HGDI indices of 0.85 and 0.82, respectively. It is followed by gki, ddl, rec and tpi with HGDI indices of 0.79; 0.57 and 0.55. The least polymorphic was yqiL with the HGDI index of 0.30. Using the eBURST algorythm, we have defined one group of three allelic profiles: ST1212, ST1224 and ST1228. No significant regional grouping of allelic profiles was visible from the MST view. Nevertheless, it should be noted that we conducted the study on 31 samples of which the vast majority have newly-identified STs. Allelic profiles are original due to new alleles at the arcC locus. Conclusion: In conclusion, the concordance of the conventional microbiological identification to the genus level with molecular identification was high (87%). The identification of the species by conventional microbiological procedures mostly coincides with S. uberis molecular identification and least for S. agalactiae. The research paves the way for further molecular epidemiology studies of streptococcal mastitis pathogens, because for some streptococcus strains laboratory procedures should be extended to determine their zoonotic nature and potential.