Bacteria are immobilized in cheese after the coagulation step whatever the technology. They grow then as bacterialcolonies whose spatial distribution (the size of colonies and the distance between them) is dependent on theirinoculation level. We showed in model cheeses, using fluorescent lactococci strains and confocal microscopy, that forthe same final biomass, the size of bacterial colonies was increased when the inoculation rate was decreased. The meandistance between bacterial colonies was 250 ?m when they were inoculated at 104 cfu/g while the mean distance was25 ?m when they were inoculated at 107 cfu/g. Consequently, the smaller and the most numerous are the colonies,the larger was the exchange surface between the bacterial colonies and the cheese matrix, so called interfacial area. Theinterfacial area was 7-fold larger when the inoculation rate was 50 times higher. This phenomenon has been validatedin real semi-hard cheeses. We also demonstrated that the composition of the cheese matrix modified the shape of thebacterial colonies, and consequently modified the interfacial area, as spherical colonies have smaller interfacial area thanthe irregular ones.We assumed that the different interfacial areas due either to different spatial distributions of the bacterial coloniesor to different cheese compositions, may lead to different micro-environments around and between colonies. Thesedifferent micro-environments may then impact the bacterial expression during ripening. The decrease of pH has beenmeasured around colonies of different sizes, during acidification, in situ using a pH-sensitive fluorescent probe (Snarf-4F). The acidification kinetics was dependant on the colony size, but micro-gradients of pH could not be evidencedaround colonies. Further explorations of the micro-environment (redox, ...) of bacterial colonies depending on theirdistribution are ongoing, and will be presented.