Stem cells confer to adult tissues the capacity to maintain morphology and function counterbalancing intrinsic wear and tear (homeostasis) and environmental damages (regeneration). Stem cells are undifferentiated multipotent cells that perpetuate themselves indefinitely. To sustain cell demand from the tissue, they generate progenitors which are able to differentiate to substitute old and/or damaged cells. In the process of cell replacement, feedback mechanisms from the tissue to stem cells ensure adequate proliferation rate and exact lineage specification. Unfortunately, cell turnover is difficult to monitor and regeneration paradigms have been widely used to infer molecular mechanisms behind homeostatic cell turnover. But actually, very little is known about how mechanisms of regeneration resemble or differ from homeostatic tissue maintenance. In this work, using Drosophila midgut as a model, we investigated directly the basic mechanisms of epithelial homeostasis, in unchallenged conditions. To this end, we devised an original method which allowed detecting tissue turnover in a precise and temporally controllable manner in midguts. We found that in normal homeostatic conditions, midgut turnover follows unexpected asynchronous dynamics and that, surprisingly, progenitor cells sense where exactly to differentiate independently of birth time. We have identified Escargot, a Snail family gene, and the miR-200-related microRNA miR-8, as key intrinsic elements controlling this progenitor behavior during homeostasis. The escargot gene hold progenitors in undifferentiated state repressing mir-8 locus and conferring them marked mesenchymal traits that we found to be a prerequisite for proper intercalation into the epithelium. Conversely, miR-8 controls through direct targeting of escargot mRNA, the transition from undifferentiated toward differentiated state by repressing mesenchymal characteristics. The break of this reciprocal regulation impacts on homeostasis by altering the spatial and timing control of progenitors differentiation. Altogether, these results indicated that progenitors are not simple transient and passive entities but active players in homeostasis. Possibly, progenitors are able to integrate local feed-back signals to regulate their cellular state via the antagonic Escargot/miR-8 action. Given the striking analogies between Escargot/miR-8 "undifferentiated to differentiated transition" and the Snail/miR-200 mesenchymal to epithelial transition (MET), we think the future identification of the signals and molecular mechanisms controlling Escargot and miR-8 would be of wide-ranging relevance to understand not only homeostasis but also physiological and pathological related MET, as wound healing and metastasis establishment.