In Sicily, the ovine livestock production has strong traditional connotations, with an almost exclusively extensively character. Moreover, the dairy branch is an important part of the ovine sector and shows several critical issues. Unfortunately, we have little informations about the hygienic sanitary quality of ovine dairy products. The aim of the study was to perform a microbiological characterization of ovine milk, ricotta and tuma from different Sicilian districts, using rapid and high reliability technologies, in order to collect data on the microbiological quality of local products. For this purpose, the prevalence of isolated strains belonging to the Family of Enterobacteriaceae has been evaluated, along with their identification with mass spectrometry techniques. Finally, all identified strains were tested for the phenotypic expression of ESBL (Extended spectrum ?-lactamase) production. Then, the ESBL positive strains were selected for the elaboration of an experimental model. This work lead to the implementation of the mass spectra database with Superspectra highly specific for the phenotypic expression of ESBL. Data collected during the study revealed poor hygienic-sanitary conditions of the dairyproducts. Milk samples showed a mean value of 6.4±1.4 Log CFU/ml for Aerobic mesophilic count and 3.8±1.8 CFU/ml for Enterobacteriaceae count. Same considerations can be made for ricotta and tuma samples, in which, in addition to those two parameters, also Staphylococcal charges resulted unsatisfactory, with a mean value of 3.9±1.3 Log CFU/g and 3.8±1.5 Log CFU/g respectively. In one sample of ricotta was also detected Listeria monocytogenes with a charge of 3.3 Log CFU/g. Among the isolated strains belonging to the family of Enterobacteriaceae (n.355), Hafnia alvei reported the highest prevalence in all the samples (40.3%), followed by E. coli (13.2%), Citrobacter spp (10.1%) and other enterobacteria in lower percentage. Considering the phenotypical exhibition of antibiotic resistance, the 70% of strains resulted ESBL+, with different expression rates depending on the species. Finally, 5 samples of ESBL + and ESBL - strains for species (Hafnia alvei, E. coli, E. cloacae, K. oxytoca) were selected in order to generate mass spectra that were analyzed for peaks similarities. The results of this comparing algorithm have conducted to a new hierarchical clustering of the samples and to the implementation of the data set with new Superspectra ESBL+ and ESBL-, that allowed both the rapid identification at the genus level and the ESBL expression of each sample. The reliability of our model ranged from 50% to 80% of correct identifications; so mass spectrometry could be routinely employed in microbiological diagnostics as novel, time-saving and cost-reducing tool.