Abstract: A set of tractable quantum mechanical models are developed and applied for the atomistic treatment of electron spin-dependent properties in colloidal semiconductor nanocrystals and fullerene-based molecules. The spin properties related to II-VI semiconductor nanocrystals are described first. The Landé g factor, and in general, the magnetic field response of electrons and excitons, is examined with a perturbative theory of the spin-orbit and magnetic field effects, to describe the qualitative effects of aspect ratio. Subsequently, a non-perturbative treatment of finite magnetic field, spin-orbit coupling, and excitonic interactions, performed using a restricted two-particle basis, single-excitation configuration interaction method is presented. Next, the magnetic properties of nanocrystals doped with small numbers of transition metal atoms are examined using a statistical mechanical treatment of the Heisenberg exchange model, and an inhomogeneous doping model is proposed to explain current experimental results. Finally, the spin-coherent transfer of electrons between nanocrystals linked by small organic molecules is examined using a Green's function transport formalism. An experimental test using quantum interference is proposed to validate the coherent transfer hypothesis. The remaining work examines the spin properties of fullerene and molecular systems. A modified Thomas-Fermi model is described for the treatment of the hyperfine coupling constant of nitrogen atoms endohedrally enclosed in fullerenes, and is used to study the comparative effects of fullerene charge and size on the hyperfine coupling constant. Next, the coherent spin-transport properties of three nanometer scale systems are discussed. A density-functional study of the gadolinium atom endohedrally enclosed in the C₈₂ fullerene, Gd@C₈₂, predicts a lack of spin-dependent transport properties, despite the large magnetic moment on the Gd atom. This is followed by a proposal, and test using density-functional calculations, of a model organometallic molecule to control spin- and charge-currents in molecular electronics devices. This proposal has the practical advantage of requiring magnetic manipulation of only a single vanadium atom, instead of magnetization reversal of a bulk metal contact, as in existing schemes. Third, a proposal is presented for the use of (5,5)@(10,10) multiwall carbon nanotubes as a nanoscale version of a slide potentiometer, and studied using a pi-orbital model Hamiltonian.