Background Goats homozygous for a null allele (O/O) at the CSN1S1 locus encoding αs1-casein display a chronic endoplasmic reticulum (ER) stress (Unfolded Protein Response) due to an accumulation of the other caseins in this compartment, thus triggering a general MEC dysfunction with a strong impact on milk composition. Milk-derived extracellular vesicles (MEVs) contain molecular information, which are thus recognized as mediators of intercellular communication. We made the assumption that the absence of CSN1S1 expression may influence MEVs cargos including miRNA, proteins, lipids and metabolites. The objective of this study was to assess the impact of αs1-casein deficit on MEVs cargos and to compare the biological material they convey. Material & Methods We have developed an improved method based on a density gradient ultracentrifugation to isolate MEVs. The quality of MEVs was analyzed morphologically by transmission electron microscopy (TEM) with negative staining (uranyl acetate), the specific “exosome” protein markers were detected by Western blot and ELISA and the size distribution and particle concentration were measured by NTA. The MEV’s proteome was acquired by LC-MS/MS and nucleic acid content (mRNA and miRNA) by NGS and qPCR. EV lipid content was extracted with the MTBE method. Data acquisition was performed by an Orbitrap-MS and analyzed using Lipid Data Analyzer. MEV metabolites were extracted using MPLEx protocol, data acquired by LC coupled with HRMS and analyzed using Metaboanalyst online workbenches. Differential statistical and bioinformatic analyses were performed using appropriate softwares. Results The novel purification method gives MEV populations, free of contamination by other EVs and milk components, at sufficient concentrations to perform subsequent analyses. Nearly 280 proteins involved in the biogenesis of exosomes and MVB formation, their adhesion and internalization as well as proteins associated with membrane transport and enzymes involved in cellular metabolism were identified, among which 41 exosomal proteins differed between CSN1S1 O/O (null) and A/A (wildtype) genotypes. Ongoing profiling of RNA from MEVs has already identified over 230 miRNA and confirmed MEC origin due to the presence of mRNA encoding specific major milk proteins. The comparison of exosomal miRNomes of goat homozygous for A and O alleles at the CSN1S1 locus pointed out 15 miRNAs differentially abundant, potentially related to the MEC phenotype. Sphingomyelin and phosphatidylcholin were the major phospholipids observed in MEV populations. We have totally identified ca. 4,000 compounds using pHILIC and RPLC, 79 of which were significantly up or down-regulated in studied genotypes. Conclusion Several differences distinguishing goats according to the genotype at the CSN1S1 locus were found at each level of the omic analysis of MEVs. Differentially abundant miRNAs and transcriptome analyses are in agreement with UPR phenotype and confirmed their involvement in post-transcriptional regulatory mechanisms.