Extracellular vesicles (EVs) are lipid membrane vesicles released by all human cells and are widely recognized to be involved in many cellular processes, both in physiological and pathological conditions. They are mediators of cell-cell communication, at both paracrine and systemic levels, and therefore they are active players in cell differentiation, tissue homeostasis, and organ remodeling. Due to their ability to serve as a cargo for proteins, lipids, and nucleic acids, which often reflects the cellular source, they should be considered the future of the natural nanodelivery of bio-compounds. To date, natural nanovesicles, such as exosomes, have been shown to represent a source of disease biomarkers and have high potential benefits in regenerative medicine. Indeed, they deliver both chemical and bio-molecules in a way that within exosomes drugs are more effective that in their exosome-free form. Thus, to date, we know that exosomes are shuttle disease biomarkers and probably the most effective way to deliver therapeutic molecules within target cells. However, we do not know exactly which exosomes may be used in therapy in avoiding side effects as well. In regenerative medicine, it will be ideal to use autologous exosomes, but it seems not ideal to use plasma-derived exosomes, as they may contain potentially dangerous molecules. Here, we want to present and discuss a contradictory relatively unmet issue that is the lack of a general agreement on the choice for the source of extracellular vesicles for therapeutic use.