In recent years, collaborative studies that leverage experimental synthesis, spectroscopic characterization, and electronic structure analysis have enabled significant advances in f-element chemistry through the discovery of newly accessible metal oxidation states and novel electronic configurations for lanthanide (Ln) and actinide (An)-containing species. With the purpose of extending the utility of computational approaches to maintain this positive trajectory, the present thesis discusses recent developments and applications of computational chemistry methods, with a focus on density functional theory (DFT), towards the accurate prediction and characterization of new f-element complexes. A DFT methodology based on (meta)-generalized gradient approximation (mGGA) density functionals, triple-ζ quality basis sets for metal atoms, and effective core potentials (ECPs) is shown to provide the electronic structure insights necessary to understand the unexpected stability of fully linear Dy and Tb-based metallocene species, Dy(CpiPr5)2 and Tb(CpiPr5)2. Calculations reveal that such unorthodox molecular geometry, which is rarely observed for lanthanides, is facilitated by a 4fn 5d1 electronic configuration of the metal center, which gives rise to a σ−bonding interaction between the 5d/6s HOMO and cyclopentadienyl ligand system. Further calculations using this DFT methodology predict the existence of stable, linear An-based metallocenes, and the results of this study are used to guide synthetic efforts towards the experimental isolation of U(CpiPr5)2, the first An-based “ferrocene” analog. A similar computational methodology which replaces the ECP with an all-electron approach is applied towards the characterization of Ln-based spin molecular qubits, [Lu(OAr*)3]−, [La(OAr*)3]−, and [Lu(NR2)3]−, revealing the importance of 6s orbital contributions to the spin density to facilitate large Fermi-contact and thus hyperfine interactions. This thesis concludes by describing the prediction of accurate EPR parameters, such as the hyperfine coupling constant, electronic g-tensor, and quadrupole coupling constant in relativistic DFT for the broader study of candidate molecular qubits. An implementation of these quantities is presented within the relativistic exact two-component theory (X2C) method, and benchmark calculations on transition metal and f-element complexes are provided to evaluate choice of the relativistic Hamiltonian, basis set, and density functional approximation (DFA). A recommended set of parameters based on the results of these benchmarks is presented, and subsequently used to calculate the EPR parameters for the previous series of Lu and La molecular qubit systems. These predictions are found to reduce errors by roughly one order of magnitude when compared with the unrefined methodology, and are highly accurate when compared to experimental data - representing an advance for in-silico characterization of EPR spectra. Present challenges and future directions for the development of electronic structure methods for the study of the f-elements are assessed.