Small RNAs are key post-transcriptional regulators of eukaryotic gene expression. Among the most fascinating aspects of small RNAs is their ability to cross cell boundaries owing to their non-cell-autonomy. Recently, the host laboratory demonstrated that 21-24bp siRNAs could act as mobile silencing signals in Arabidopsis. Interestingly, in C. Elegans, systemic silencing requires SID1, a transmembrane channel through which endogenous double-stranded RNA may be communicated to adjacent cells. Functional SID1 homologues and miRNAs found in secreted exosomes in mammals suggest that systemic RNA silencing might also operate in these organisms, raising the question of how this process might be regulated? Tight regulation is indeed anticipated given the exquisite expression patterns and developmental roles of many mammalian miRNAs. A first possibility for regulated miRNA movement entails that it might mostly occur between compatible “emitting” and “receiving” cells. This might be achieved via qualitatively differences in miRNAs effector complexes, localization or shear availability of silencing transporter systems. A second, non-mutually exclusive possibility is that release of miRNA through membranes might be polarized. The identification, in the host laboratory, of a requirement for multi-vesicular bodies for the assembly of miRNA effector complexes supports this idea. In fact, we propose that both mechanisms could be at work in specialized cells, such as secreting epithelia, to direct the selective release of miRNAs either along the epithelial cell layer or in body fluids. Using the mammary gland as a model system we will (i) decipher the molecular requirements for cell to “emit” or “receive” systemic miRNAs (ii) study how cell-polarization might affect miRNA cell-to-cell transfer (iii) investigate the in vivo relevance of our findings by characterizing miRNAs contained in milk, and by studying potential effects of systemic miRNA arising from grafted tumors in mice.