The Kohn-Sham formulation of density functional theory (DFT) has posed itself as one of the most popular and versatile methods for condensed phase studies owing to its reasonable accuracy and affordable computational cost. DFT, in principle, yields exact ground state energy, including dispersion forces that are of primordial importance in chemical and biological systems. Yet with many exchange-correlation functionals in practical use such as the local density approximation or generalized gradient approximations, DFT either provides sporadic results or fails completely to account for these forces. In consequence, various methods offering remedy for this shortcoming have been proposed in this active field of research. In particular, dispersion-corrected atom-centered potentials (DCACPs) serve as a robust and efficient way to include these weak forces in a fully self-consistent manner within current DFT frameworks. The aim of this thesis is twofold: first, to improve the predictive power and the understanding of the DCACP concept; second, applying DCACPs to systems of increasing complexity starting with dimers, continuing through larger clusters and ending with the condensed phase. The success of the second aim not only justifies the use of DCACPs but more importantly, provides insights to the role dispersion forces play in the systems investigated. We first draw on the atoms-in-molecules theory and a multi-center density expansion to justify the form and universality of DCACPs. A library of DCACPs calibrated with an improved penalty functional against high-level ab initio references is presented. With the library in hand, we extend our studies to systems of biological significance, mainly constituents of proteins and DNA; polycyclic aromatic molecules intercalated in between segments of DNA are the center of focus. The application of DCACPs is then furthered to the condensed phase and the importance of van der Waals interactions in liquid water is investigated.