Traditional plating on M17 medium led to loads ranging from 105 to 109 CFU/g, including cheese samples were no L. lactis was found by RT-qPCR. In these cheeses, none of the colonies isolated on M17 medium was identified as L. lactis. These data could be interpreted as a lack of selectivity of M17 medium where colony growth is not always related to lactococcal species. Probably, lactococci are able to grow on M17 medium when they are abundant and not stressed, as for example during milk and curd fermentation. Differently, during the ripening process, it is known that NSLAB increase in number and prevail on lactococcal populations, which are often out-competed by the numerically more abundant lactobacilli. Nevertheless, in this work, a few isolates were identified as L. lactis by His-PCR. They were obtained from eight cheese samples with loads higher than 107 CFU/g, detected by RT-qPCR, except for two samples characterized by values of 104 and 106 CFU/g. These data could be explained with the relative high abundance of L. lactis in these cheeses and, thus, its capability to compete with the rest of microbiota and multiply on synthetic media. Currently, M17 is the medium mainly used for lactococci cultivation, but new formulations for the isolation of LAB from cheese have been recently studied as, for example, cheese agar, which was used to recover minority populations from milk, whey starter and fresh curd of Parmigiano Reggiano, hardly estimable on traditional media. Thus, as future prospective, for a more reliable and effective recovery of lactococci, in particular L. lactis, during cheese ripening, the optimization and formulation of specific nutritional conditions should be better investigated. The absence or low abundance of L. lactis isolates on M17 medium, support the thesis hypothesized by other authors that L. lactis starter populations are mainly present in VNC state during cheese ripening and, for this reason, culturedependent methods are not able to detect their presence and have to be complemented with direct analysis in cheese. These considerations can be especially corroborated from the results obtained in eight cheese samples where the difference, in terms of microbial load, between RT-qPCR and plating data, was lower than 102 CFU/g and, thus, the absence of L. lactis growth, on M17, could not be justifiable with the prevalence of NSLAB. For some of the cheeses analysed, experiments were performed in order to “resuscitate” L. lactis VNC cells and preliminary results highlighted that different AG-013736 carbon sources, in cultural media, affect differently their growing ability ; in particular, enrichment in medium with high percentage of glucose seemed to stimulate the attitude of the cells to become culturable again. Recent researches were focused on these aspects and highlighted the presence of VNC L. lactis cells in ripened cheese products. Differently, Flo´rez and colleagues found abundance of L. lactis isolates on M17 from the analysis of Spanish cheese, but they did not specify the distribution of the isolates among milk, curd and cheese samples.