Aims: The microbial and chemical composition of seven different semi-ripened (45 days) Provola dei Nebrodi Sicilian cheese samples were assessed in order to investigate the diversity of the microbial population in cheese made from different geographical areas throughout Sicily. Methods and Results: The samples, which were obtained from seven different Provola dei Nebrodi manufacturers, were assessed using selective media. Interestingly, concentrations of presumptive lactobacilli represented over 90% of the total microbial population. In total, 105 presumptive Lactobacillus isolates were characterized to determine the relatedness of the isolates between the seven different cheeses. Randomly amplified polymorphic DNA polymerase chain reaction (RAPD PCR) analysis of the 105 presumptive lactobacilli indicated the presence of 22 distinct isolates. Further investigation of the isolates using pulsed field gel electrophoresis (PFGE) following restriction with the enzyme ApaI revealed the presence of 19 distinct macrorestriction patterns and the presence of between one and four distinct isolates per cheese sample (out of a total of 15 isolates per cheese randomly taken from Lactobacillus selective media plates). Analysis of the 16S rDNA sequence of each genetically distinct isolate demonstrated the dominance of the Lactobacillus casei species in all cheese samples assessed. Lactobacillus delbrueckii and Pediococcus pentosaceus species were also detected. The concentration of free amino acids, used to estimate the extent of proteolysis in each cheese, ranged from 59 to 433 mg 100 g(-1) cheese. Conclusions: Microbiological assessment of the cheeses demonstrated the dominance of Lactobacillus species after 45 days of ripening with levels ranging from 8.3 to 9.4 log CFU g(-1). Significance and Impact of the Study: This study provides new information on the diversity of lactobacilli within an artisanal Sicilian cheese, enabling the identification of 17 strains of Lact. casei, one strain of Lact. delbrueckii and Ped. pentosaceus through the combined use of RAPD PCR, PFGE and 16S rDNA sequencing.