Nanomedicine is promising to improve conventional cancer medicine by making diagnosis and therapy more accurate and more effective in a more personalized manner. A key of the cancer nanomedicine is construction of medical nanodevices by programming various requisite functions to nanoparticles (NPs). As compared to that of soft NPs, including organic micelles and polymers, fabrication of an inorganic NP based nanodevice is still challenging; the approved nanoformulations have been confined to the limited number of superparamagnetic iron oxide NPs (SPIONs). The major challenges lie in how to program the requisite functions to inorganic NPs. In spite the much denser and less hydrophilic properties of inorganic NPs, most of the following functions have to be programmed for their in vivo applications: (A) high dispersibility in a physiological environment, (B) high stealth efficiency to slip through the trap by liver and spleen, (C) high targeting efficiency to cancer tissue, (D) clear visualization of cancer for diagnosis, and (E) high anticancer activity for treatment.In our approach, poly(glycerol) (PG), containing a hydroxy group at every monomer unit, was found as a better alternative to poly(ethylene glycol) (PEG), the most commonly used hydrophilic polymer, giving (A) high dispersibility to inorganic NPs. Although most of the inorganic NPs are not dense in functional groups, the hyperbranched structure with many hydroxy groups in PG turns the less functional surface into highly functional one, imparting not only good hydrophilicity but also (B) high stealth efficiency as we reported recently. In addition, a number of hydroxy groups in PG afford the structural or functional extensibility to introduce the additional layer or function. This enables us to design and construct a three-layer architecture consisting of a core inorganic NP, a hydrophilic and stealthy PG layer, and a functional molecule layer, where their interfaces are connected firmly by covalent bonds. The three-layered nanodevice is very flexible in its design for the following reasons: The PG coating can be applied to a wide variety of inorganic NPs with various functions, and various functional moieties can be introduced on the PG layer as a functional molecule layer. Owing to the versatility of the three-layer model, the rest of the above functions (C)-(E) can be programed in the NP core and/or the outmost layer in nanodevices.In this Account, the author described first the methodology for precise construction and quantitative characterization of various biomedical nanodevices. This fundamental aspect of this research has been achieved by "applying organic chemistry to nanomaterials" which is the concept of our research. That is, the rich chemistry in synthesis and characterization of organic compounds has been applied to the nanodevice fabrication and characterization. Second, evaluation of the functions programmed in the nanodevices is described in terms of stealth and targeting efficiencies, cancer diagnosis and therapy, and biomedical sensing. This stage in our research made us more interdisciplinary from chemistry and nanoscience to biology and medicine. The following research spiral has been established in our group to strongly promote the improvement of our biomedical nanodevices; nanodevice design → precise construction → quantitative characterization → functional evaluation.